Problem Set 3 This is the final homework assignment, which accounts for of your final grade. Unlike the previous problem sets, you are required to collect the data on your own and conduct data analysis based on your collected data. You may work with other students. The maximum number of students per group is two. However, you can work on your own. Be sure to indicate with whom you have worked in your submission. Deadline: Dec 5, 2024 (HK Time 11:59 PM). There is a penalty for late submissions: will be subtracted from the total mark for every additional day after the deadline. If you submit it after Dec 15, 2024, you will get a zero on this homework assignment. Reference Ang, A., Hodrick, R. J., Xing, Y., & Zhang, X. (2006). The cross-section of volatility and expected returns. Journal of Finance, 61(1), 259-299. Background In this problem set, you will examine the pricing of volatility risk in the cross-section of stock returns, following the Journal of Finance paper Ang, Hodrick, Xing, and Zhang (2006) (thereafter, AHXZ (2006)). Specifically, we ask the following questions: Do the stocks with larger exposures to the volatility risk earn higher or lower average returns? To answer this question, we first need to find a measure of volatility exposure. Following AHXZ (2006), we consider the VIX index. The VIX index is constructed so that it represents the implied volatility of a synthetic at-the-money option contract on the S&P100 index that has a maturity of 1 month. It is constructed from eight S&P100 index puts and calls and takes into account the American features of the options contracts, discrete cash dividends, and microstructure frictions such as bid-ask spreads. Because the VIX index is highly serially correlated with a first-order autocorrelation of 0.94, we measure daily innovations in aggregate volatility by using daily changes in VIX, which we denote as ΔVIX. There are three parts in this problem set. Part I. Main Findings of AHXZ (2006) In this part, you need to summarise the main findings in AHXZ (2006). Please list the two findings that you think are the most important (there are more than two, but you do not have to list all of them). For each key finding, please provide an economic explanation of the empirical phenomenon. Part II. Collecting Data The monthly and daily individual stock data come from CRSP, accounting data come from COMPUSTAT, and data on the CBOE implied volatility index, VIX, come from the FRED St Louis. To be clear, AHXZ (2006) use the SP100-based implied volatility index, which has a ticker of VXO, for all tests reported in this paper. You can download the Fama-French three factors (market, size, and value factors) from Ken French's website. I provide some useful links to several datasets at the end of this document. Your first task is to download all the data and load the datasets using pandas . After that, you need to report (1) which datasets you use in this problem set and why you need them, (2) how you preprocess the data (e.g., dropping samples based on some requirements, handling missing data, merging datasets, etc.), and (3) how many firms per year your final sample has in the panel data of stock returns. Part III. Pricing Aggregate Volatility Shocks To measure the sensitivity to aggregate volatility innovations, you are required to run the following regression: where: MKT-RF is the daily market excess return, ΔVIXt is the daily change in the VIX index, and βiMKT and βiΔVIX are firm i's loadings on market risk and aggregate volatility risk, respectively. You need to run the above regression with daily data for each stock per month. Specifically, for each month, you run the regression for all stocks on AMEX, NASDAQ, and the NYSE, with more than 17 daily observations and obtain the monthly estimates of βiMKT and βiΔVIX. In this step, you will need to use the CRSP daily stock return data and also the market and VIX daily data. At the end of each month, you sort stocks into quintiles based on the value of the realized βiΔVIX coefficients over the past month. Firms in quintile 1 have the lowest coefficients, while firms in quintile 5 have the highest loadings. Within each quintile portfolio, we value-weight the stocks. We link the returns across time to form. one series of post-ranking returns for each quintile portfolio. In this portfolio sorting step, you should use the CRSP monthly stock return data. Your task is to replicate the empirical results in Table I of AHXZ (2006). Please only replicate the numbers in the following attached table. The first two columns report the mean and standard deviation of the monthly total, not excess, simple returns. The column labelled % Mkt share shows the percentage of market cap for all the stocks in each quintile. The columns labelled size and B/M show the average log market capitalization and book-to-market ratio for firms within the portfolio (You do NOT need to replicate these two columns). The columns labelled “CAPM Alpha” and “FF-3 Alpha” report the time-series alphas of these portfolios relative to the CAPM and to the FF-3 model, respectively. The final column reports the pre-formation βiΔVIX coefficients, which are computed at the beginning of each month for each portfolio and are value-weighted. The sample period in AHXZ (2006) is from January 1986 to December 2000. However, I require you to conduct the same data analysis in the out-of-sample, January 2001 to December 2020. Does the long-short portfolio (marked as 5-1 above) have similar performance in the more recent sample from January 2001 to December 2020? How do you interpret your findings? Caveat: It is impossible to get exactly the same numbers as in the original paper. Some useful links: VXO/VIX index: https://wrds-www.wharton.upenn.edu/pages/get-data/cboe-indexes/cboe-indexes-1/cboe-indexes/ or https://fred.stlouisfed.org/series/VXOCLS Ken French's library: https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html CRSP Stock / Security Files: https://wrds-www.wharton.upenn.edu/pages/get-data/center-research-security-prices-crsp/annual-update/stock-security-files/ To save you time, I put the daily and monthly stock return data in Dropbox. Please refer to datasets_ps3.ipynb posted on Moodle. However, you need to download the VXO index and Fama-French factors on your own, using the above links. Submission Requirement You need to submit two documents: A PDF file that contains your explanation, such as the main findings, economic explanations, execution details, etc. Please keep your document as concise as possible (no more than five pages). Python codes (Jupyter Notebook, .py file, etc.) that show the details of your data work (Please add as many comments as you can). Your grade is determined by the accuracy of your solutions, explanations of each data analysis step, and your interpretation of the empirical findings. Please do NOT submit the datasets. Finally, if you find anything unclear, please read the JF paper, AHXZ (2006), carefully.
FIN212A- FINANCIAL STATEMENT ANALYSIS PROJECT OVERVIEW The primary purpose of this project is to provide you with an in-depth look at audited financial statements. You will be conducting a comprehensive analysis of the financial statements and related notes contained in the audit reports of two organizations that you choose to study selected from a list of paired companies. The specific steps that you should follow in completing this project are presented below. PROJECT LOGISTICS The project is to be conducted on a group basis. Each student will be assigned a group by the professor. Groups will include five students. Each group is responsible for choosing, in priority of preference, three sets of paired companies from a list at the bottom of this document. Every effort will be made to assign a pair from your preference list but there is no promise that we will be able to accommodate all requests. If you prefer to study another pair of public companies that do not appear on the list, please check with me to determine if these companies are a good fit for the project. To complete the project, you need to obtain audited financial statements for the most recent fiscal year for these companies. These may be found in the investor relations section of the company’s website. It can also be accessed through the SEC and financial aggregators. Students are encouraged to go to the Business and Economics Library page for other sources of financial information, such as FactSet and CapitalIQ. Each group is expected to electronically submit two files reflective of their group’s effort. The first is an excel file which includes a list of the following required exhibits for each of the two companies in the analysis: 1. Comparative annual balance sheets for the most current year end, and the previous three year-ends. 2. Comparative annual income statements for the most current year, and the previous three years. 3. Comparative annual statement of cash flows, for the most current year, and the previous three years. 4. Most current annual statement of stockholders’ equity. 5. Common-size balance sheets and income statements for the most current year and the previous three years (for balance sheets choose total assets as the base item, for income statements choose total sales as the base item) 6. Extended DuPont Analysis (decompose both ROE and ROA). Be sure to show each element of the formula 7. Industry Benchmarks for DuPont Analysis components for the most recent year and the previous year 8. At least five additional ratios that your group has chosen to calculate based on the questions generated in your DuPont Analysis (make sure to include at least one activity ratio, one liquidity ratio, one solvency ratio and one market test ratio) Note: Common-size statements, as well as ratios, should be clearly presented and should clearly show the formulas used in the calculation. In each of the required exhibits, make sure the most recent year appears on the left and the previous years on the right. The second file to be submitted on behalf of the group is a Word file which should not exceed 10 double spaced pages with font size 12. Each report should start with a cover page, which is not included in the 10-page count. On the cover page you should provide a table that summarizes each member’s contribution in percentage to the overall project (the total should equal 100%). Please be very careful about citing sources where appropriate. Each report should conclude with a bibliography. Appendix with exhibits should be included after the bibliography section. The tables and figures presented in these exhibits should be numbered and referred to within the 10-page analysis. Make sure to fit teach table or figure neatly into one page, and list the source in a footnote. PROJECT REQUIREMENTS (Your analysis should include the following but is not limited to these items.) 1. The project should begin with an introduction describing the nature of the entities being analyzed: What industry are the two companies in? What are the most important characteristics of the industry? What are the opportunities and challenges the industry faces in the current and future environment? Describe each of the two companies. Include a description of their products, markets, competitors and the opportunities and challenges each of them faces. 2. Review the common-size income statements and balance sheets for each company, identify and discuss significant changes that warrant attention (a change of 2% or more is considered significant). Investigate and provide explanation for these changes. 3. Ratio Analysis: I. Using the Extended DuPont analysis as reference: a) Explain what each ratio means in the context of the companies that you are analyzing (do not copy the definition from the textbook. Use your own words) b) Identify significant changes over the three-year analysis period. c) Explain the reason(s) for these changes based on information in the notes to the financial statements and in the Management Discussion and Analysis. If necessary, you may refer to outside reports and analysis including stock analyst reports. Indicate whether you believe that these are favorable or unfavorable trends and discuss your interpretation of such trends along with their implications. d) Identify which company is stronger in each component of DuPont Analysis and also determine whether each company is performing above, below, or at industry performance levels. II. Based on the above analysis, choose at least five but no more than seven additional ratios (at least one activity ratio, one liquidity ratio, one solvency ratio and one market test ratio) which allow you to analyze more deeply the strengths and/or weaknesses you have identified in your DuPont analysis. Repeat steps (a) through (d) in part I. Discuss what these ratios add to your understanding of the companies’ financial performance and what they indicate about operations. 4. Use the 10-K to Obtain answers for the following questions: a) What type of inventory cost flow assumption(s) does each company use? (Do not copy the paragraph from the 10-K. Provide only one line for each company). If one or both of the companies uses LIFO, explain what would have been the impact of using FIFO on the comparison of the inventory turnover. b) What type of depreciation methods are employed by the companies that you are examining? (Do not copy the paragraph from the 10-K. Provide one or two lines for each company). c) Are there liabilities that may exceed the amounts reflected in each company’s balance sheet? Refer to the note on contingent liabilities, as well as any notes on off-balance sheet items, and explain what would have been the impact of these amounts had they been recorded in the balance sheet. Clearly specify the source for this information. d) Review the statement of stockholders’ equity for each of the two companies. Identify the major transactions that impacted the stockholders’ equity in the current year and the previous one (Do not copy the paragraph from the 10-K. Use your own words). e) Review the statement of cash flows for each company and identify their major sources and uses of cash. (Do not copy the paragraph from the Management Discussion and Analysis. Use your own words) f) Read the Reports of the Independent Auditors for each of the companies for the most current year. Are these reports reflective of standard unqualified audit reports? If not, explain the departure from the standard and indicate whether or not the departure is favorable or unfavorable. (Do not copy the paragraph from the 10-K. Use your own words, and provide only one line for each company if the report is reflective of standard unqualified audit report). 5. Finally, if you had $500,000 to invest in one of the two companies you have analyzed, which would be the better investment and why? Make sure to utilize the analysis you have done earlier in the paper to support this decision. In addition, read and reference at least one article on the industry and one stock analyst report on each company. (If you studied valuation in prior course work, you may comment on the expected value of the investment at the end of 5 years, but this is not required) In addition to the Excel file and the Word file which are to be submitted once on behalf of all group members, each group member is required to submit a confidential peer evaluation, where you will be asked to rate each other’s contributions, team work, responsiveness and communication. These peer evaluations will be taken into account for grading purposes. Pairs of Companies for Analysis: 1. Technology: 11. Semiconductors: 21. Retail - Department Stores: Microsoft Corporation (MSFT) Intel Corporation (INTC) Macy's, Inc. (M) Apple Inc. (AAPL) Advanced Micro Devices, Inc. (AMD) Nordstrom, Inc. (JWN) 2. Automotive: 12. Food & Beverage: 22. Retail - Home Improvement: Ford Motor Company (F) The Coca-Cola Company (KO) The Home Depot, Inc. (HD) General Motors Company (GM) PepsiCo, Inc. (PEP) Lowe's Companies, Inc. (LOW) 3. Software: 13. Insurance: 23. Retail - Discount Stores: Salesforce.com, Inc. (CRM) The Allstate Corporation (ALL) Dollar General Corporation (DG) ServiceNow, Inc. (NOW) Progressive Corporation (PGR) Dollar Tree, Inc. (DLTR) 4. Airlines: 14. Apparel: 24. Retail - Specialty Stores: Delta Air Lines, Inc. (DAL) Nike, Inc. (NKE) Best Buy Co., Inc. (BBY) American Airlines Group Inc. (AAL) Under Armour, Inc. (UA) GameStop Corp. (GME) 5. Telecommunications: 15. Healthcare: 25. Retail - Apparel: AT&T Inc. (T) UnitedHealth Group Incorporated (UNH) The Gap, Inc. (GPS) Verizon Communications Inc. (VZ) Cigna Group (CI) American Eagle Outfitters, Inc. (AEO) 6. Pharmaceuticals: 16. Entertainment: 26. Retail - Beauty: Pfizer Inc. (PFE) The Walt Disney Company (DIS) Ulta Beauty, Inc. (ULTA) Merck & Co., Inc. (MRK) Comcast Corporation (CMCSA) Sally Beauty Holdings, Inc. (SBH) 7. Banking: 17. Real Estate: 27. Retail - Sporting Goods: JPMorgan Chase & Co. (JPM) Zillow Group, Inc. (ZG) Dick's Sporting Goods, Inc. (DKS) Bank of America Corporation (BAC) Redfin Corporation (RDFN) Foot Locker (FL) 8. Energy: 18. Biotechnology: 28. Retail - Grocery: Exxon Mobil Corporation (XOM) Amgen Inc. (AMGN) Kroger Co. (KR) Chevron Corporation (CVX) Gilead Sciences, Inc. (GILD) Albertsons Companies, Inc. (ACI) 9. Consumer Goods: 19. Industrial: 29. Retail - Home Furnishings: Procter & Gamble Co. (PG) Caterpillar Inc. (CAT) Wayfair (W) Colgate-Palmolive Company (CL) Deere & Company (DE) Williams-Sonoma, Inc. (WSM) 10. E-commerce: 20. Retail: 30. Retail - Pet Supplies: Amazon.com, Inc. (AMZN) Walmart Inc. (WMT) Petco Health and Wellness (WOOF) eBay Inc. (EBAY) Target Corporation (TGT) Chewy, Inc. (CHWY)
Chemistry 125-225: Machine Learning in Chemistry Fall Quarter, 2024 Homework Assignment #3 - Due: December 5, 2024. Turn in a writeup with your responses to a ll questions below, codes, outputs (e.g. graphs, etc.). Attach all your Python files as well, so we can run them. Problem 1: K-means Clustering on a Chemistry Dataset In this problem, you will explore k-means clustering on a chemical dataset using the scikit-learn library. The dataset contains molecular descriptors and water solubility values for various compounds. Your tasks include applying clustering, interpreting results, and evaluating the performance of your clustering model. Dataset: You will use the delaney solubility with descriptors dataset available through the DeepChem library. 1. Loading and Exploring the Data (a) Download and load the dataset using DeepChem. Convert it into a Pandas DataFrame with feature columns and a target column for solubility values. The following code snippet will help you get started: import deepchem as dc import pandas as pd # Load the dataset tasks , datasets , transformers = dc . molnet . load_delaney ( featurizer =’ECFP ’, splitter = None ) dataset = datasets [0] # Convert to Pandas DataFrame. X = dataset .X y = dataset .y columns = [f’feature_ {i}’ for i in range (X. shape [1]) ] df = pd . DataFrame. (X , columns = columns ) df [’solubility ’] = y # Display basic information about the dataset print ( df . head () ) print ( df . describe () ) (b) Describe the data briefly. How many samples and features are there? Are there any missing values? If so, how would you handle them? 2. Data Preprocessing (a) Standardize the features so that they have zero mean and unit variance using the StandardScaler class from sklearn. Explain why standardization is important for k-means clustering. from sklearn . preprocessing import StandardScaler # Standardize the features scaler = StandardScaler () X_scaled = scaler . fit_transform. ( df . drop (’solubility ’, axis =1) ) 3. Applying K-means Clustering (a) Apply k-means clustering to the standardized data with k = 3 clusters. Use the KMeans class from sklearn to fit the model and predict the cluster assignments. from sklearn . cluster import KMeans # Apply K- means clustering k = 3 kmeans = KMeans ( n_clusters =k , random_state =42) clusters = kmeans . fit_predict ( X_scaled ) df [’cluster ’] = clusters (b) Reduce the dimensionality of the data to 2 components using Principal Component Analysis (PCA) and plot the clusters. Use the PCA class from sklearn. Hint: Use matplotlib for plotting. 4. Evaluating Clustering Performance (a) Compute the silhouette score for the clustering using the silhouette score function from sklearn. What does the silhouette score indicate about the quality of the clustering? from sklearn . metrics import silhouette_score # Compute silhouette score silhouette_avg = silhouette_score ( X_scaled , clusters ) print (f" Silhouette Score for k={k}: { silhouette_avg :.3f}") (b) Experiment with different values of k (e.g., k = 2, 4, 5) and compute the silhouette score for each value. Plot the silhouette scores as a function of k to determine the optimal number of clusters. 5. Exploring Clustering Interpretability (a) Examine the cluster centroids. What patterns, if any, do you observe? Are there any features that clearly distinguish one cluster from another? (b) Select a few representative samples from each cluster (if there are at least three points per cluster) and compare their solubility values. What can you infer about the relationship between cluster membership and solubility? Use the following code snippet to sample data: # Sampling representative data points from each cluster for cluster_label in range (k) : cluster_data = df [ df [’cluster ’] == cluster_label ] num_samples = min (3 , len ( cluster_data )) # Ensure we do not sample more than available data points samples = cluster_data . sample ( num_samples , random_state =42) print (f" Cluster { cluster_label } samples :") print ( samples [[ ’solubility ’, ’cluster ’]]) 6. Advanced Analysis and Interpretations (a) Compare the performance of k-means clustering with another clustering algorithm, such as Ag- glomerative Hierarchical Clustering or DBSCAN, using the same dataset. Which method performs better based on silhouette scores, and why might this be the case? Provide code for your chosen alternative clustering method and a brief analysis of your results. (b) K-means assumes that clusters are spherical and equally sized. Discuss whether this assumption is reasonable for the dataset you are using. Based on your understanding of chemical descriptors, do you expect clusters to have such shapes? If not, suggest preprocessing or alternative methods that might improve clustering performance. (c) Use elbow method analysis to determine the optimal number of clusters for k-means. Plot the sum of squared distances (inertia) for a range of cluster values (e.g., k = 1 to k = 10) and identify the ”elbow point.” Explain your choice of the optimal k based on this analysis. # Elbow method plot inertia = [] k_range = range (1 , 11) for k in k_range : kmeans = KMeans ( n_clusters =k , random_state =42) kmeans . fit ( X_scaled ) inertia . append ( kmeans . inertia_ ) import matplotlib . pyplot as plt plt . figure ( figsize =(8 , 5) ) plt . plot ( k_range , inertia , marker =’o’) plt . title (’Elbow Method for Optimal k’) plt . xlabel (’Number of Clusters (k)’) plt . ylabel (’Sum of Squared Distances ( Inertia )’) plt . show () (d) Investigate the impact of feature selection on clustering performance. Remove features that have near-zero variance or high correlation with other features, and re-run the k-means clustering. How does feature selection impact the silhouette score and cluster interpretability? Provide a brief explanation supported by code and results. (e) Use a dimensionality reduction technique other than PCA (e.g., t-SNE) to visualize the clusters in a lower-dimensional space. Compare the visualization and interpretability of clusters using PCA versus t-SNE. Discuss any differences in the separation and distribution of clusters. Problem 2: Decision Trees and their Application in Chemistry In this problem, you will learn about decision trees, a type of supervised learning algorithm used for clas- sification and regression tasks. Decision trees model data by splitting it based on feature values, creating a tree-like structure of decisions. You will explore their application using a chemistry dataset. Dataset: You will use the delaney solubility with descriptors dataset available through the DeepChem library, which contains molecular descriptors and water solubility values. 1. Introduction to Decision Trees (a) Read about decision trees: Decision trees split data into branches based on feature values, creating a structure resembling a tree. At each node, a decision is made to split the data further based on some feature until a final prediction is made at a leaf node. Research and briefly describe: • How decision trees make predictions (e.g., how data is split at each node). • The advantages and disadvantages of decision trees compared to other algorithms. 2. Loading and Exploring the Data (a) Download and load the delaney solubility with descriptors dataset using the code below. Create a Pandas DataFrame with feature columns and a target column for solubility values. import deepchem as dc import pandas as pd # Load the dataset tasks , datasets , transformers = dc . molnet . load_delaney ( featurizer =’ECFP ’, splitter =’random ’) train_dataset , valid_dataset , test_dataset = datasets # Convert the training set to a Pandas DataFrame. X_train = train_dataset .X y_train = train_dataset .y columns = [f’feature_ {i}’ for i in range ( X_train . shape [1]) ] df_train = pd . DataFrame. ( X_train , columns = columns ) df_train [’solubility ’] = y_train (b) Describe the dataset: How many features and samples are present? What is the target variable? Print the first few rows to inspect the data. 3. Training a Decision Tree Regressor (a) Use sklearn’s DecisionTreeRegressor to fit a decision tree model to the training data. Split the features and target variable as follows: from sklearn . tree import DecisionTreeRegressor # Define features and target X = df_train . drop (’solubility ’, axis =1) y = df_train [’solubility ’] # Fit the decision tree regressor model = DecisionTreeRegressor ( random_state =42) model . fit (X , y) (b) Examine the structure of the tree using the plot tree function from sklearn. Does the tree have a large number of splits? What does this imply about the model’s complexity? from sklearn import tree import matplotlib . pyplot as plt plt . figure ( figsize =(12 , 8) ) tree . plot_tree ( model , max_depth =3 , filled = True , feature_names =X . columns ) plt . show () 4. Evaluating the Model (a) Use the trained model to make predictions on the validation data. Compute the Mean Absolute Error (MAE) and Mean Squared Error (MSE) as performance metrics. from sklearn . metrics import mean_absolute_error , mean_squared_error # Load validation data X_valid = valid_dataset .X y_valid = valid_dataset .y # Make predictions y_pred = model . predict ( X_valid ) # Calculate performance metrics mae = mean_absolute_error ( y_valid , y_pred ) mse = mean_squared_error ( y_valid , y_pred ) print (f" Mean Absolute Error ( MAE): { mae :.3 f}") print (f" Mean Squared Error ( MSE ): { mse :.3 f}") (b) Interpret the performance metrics. Is the model accurate in predicting solubility? What do the values of MAE and MSE suggest? 5. Improving the Decision Tree (a) Decision trees can easily overfit the training data. Try limiting the maximum depth of the tree (e.g., max depth=3) and re-evaluate the model using MAE and MSE. How does limiting the depth impact performance on the validation data? (b) Experiment with other hyperparameters, such as min samples split and min samples leaf. How do these parameters affect the model’s performance and complexity? Provide a brief analysis supported by code and results. 6. Advanced Analysis (Challenging Question) (a) Feature importance: Use the feature importances attribute of the trained model to identify the most important features for predicting solubility. Plot the feature importances and interpret the results. Do these features make sense in a chemical context? import matplotlib . pyplot as plt import numpy as np # Plot feature importances importance = model . feature_importances_ feature_names = X. columns indices = np . argsort ( importance ) [:: -1] plt . figure ( figsize =(10 , 6) ) plt . bar ( range ( len ( importance ) ) , importance [ indices ]) plt . xticks ( range ( len ( importance )) , feature_names [ indices ], rotation =90) plt . title (’Feature Importances ’) plt . show () (b) Discuss any limitations you observe when using decision trees for this dataset. Suggest potential approaches to overcome these limitations (e.g., using ensemble methods such as Random Forests). Problem 3: Exploring t-SNE for Dimensionality Reduction In this problem, you will learn about t-SNE (t-Distributed Stochastic Neighbor Embedding), a widely used dimensionality reduction algorithm, and apply it to visualize chemical datasets. Answer all subproblems below to demonstrate your understanding of t-SNE and its applications in machine learning for chemistry. (a) Introduction to t-SNE 1. What is t-SNE? Write a detailed explanation of t-SNE, covering: (a) Its purpose as a dimensionality reduction technique. (b) The key concepts of pairwise similarity, high-dimensional probability distributions, and low- dimensional embeddings. (c) How t-SNE minimizes the KL divergence between high-dimensional and low-dimensional distri- butions. (d) Common applicationsoft-SNE in chemistry (e.g., clustering molecular features, visualizing datasets). 2. Explain the following t-SNE parameters and their effects: (a) perplexity: How does it control the size of the neighborhood in high-dimensional space? (b) learning rate: What happens if it is set too high or too low? (c) n iter: Why is it important to use enough iterations? (d) metric: Which distance metrics can be used, and why might certain metrics be preferred in chemistry? (b) Loading and Preprocessing Data Choose a chemical dataset (e.g., ChEMBL, ESOL, or MOSES) and write Python code to: 1. Load the dataset into a Pandas DataFrame. 2. Standardize the features using StandardScaler. 3. Display the first few rows of the dataset. import pandas as pd from sklearn . preprocessing import StandardScaler # Load dataset chem_data = pd . read_csv (’ chemistry_data . csv ’) print ( chem_data . head () ) # Select numerical features and standardize features = [’ molecular_weight ’, ’alogp ’, ’hba ’, ’hbd ’, ’psa ’] X = chem_data [ features ] scaler = StandardScaler () X_scaled = scaler . fit_transform. (X) (c) Applying t-SNE 1. Use t-SNE to reduce the dataset to two dimensions using sklearn.manifold.TSNE. 2. Use the following parameter values: perplexity=30, learning rate=200, n iter=1000. 3. Write Python code to generate a scatter plot of the t-SNE embedding using matplotlib or seaborn, with points colored by a categorical property (e.g., bioactivity). from sklearn . manifold import TSNE import matplotlib . pyplot as plt import seaborn as sns # Apply t-SNE tsne = TSNE ( n_components =2 , perplexity =30 , learning_rate =200 , n_iter =1000 , random_state =42) embedding = tsne . fit_transform. ( X_scaled ) # Plot t- SNE embedding plt . figure ( figsize =(8 , 6) ) sns . scatterplot (x= embedding [: , 0] , y= embedding [: , 1] , hue = chem_data [’ bioactivity ’], palette = ’viridis ’) plt . title (’t- SNE Embedding of Chemical Dataset ’) plt . xlabel (’t- SNE 1’) plt . ylabel (’t- SNE 2’) plt . show () (d) Experimenting with Parameters Repeat the t-SNE projection with the following parameter combinations: 1. perplexity=10, learning rate=50 2. perplexity=50, learning rate=500 Plot all three embeddings side by side. Discuss how changes in perplexity and learning rate affect the embedding. (e) Comparing t-SNE with PCA 1. Apply PCA to reduce the dataset to two dimensions. 2. Generate a scatter plot of the PCA embedding. 3. Write a paragraph comparing t-SNE and PCA in terms of their ability to preserve data structure. How does t-SNE’s focus on local structure differ from PCA’s emphasis on global variance? from sklearn . decomposition import PCA # Apply PCA pca = PCA ( n_components =2) pca_embedding = pca . fit_transform. ( X_scaled ) # Plot PCA embedding plt . figure ( figsize =(8 , 6) ) sns . scatterplot (x= pca_embedding [: , 0] , y = pca_embedding [: , 1] , hue = chem_data [’bioactivity ’], palette =’viridis ’) plt . title (’PCA Embedding of Chemical Dataset ’) plt . xlabel (’PCA 1’) plt . ylabel (’PCA 2’) plt . show () (f) Reflection and Analysis Answer the following questions: 1. Do you observe distinct clusters in the t-SNE embedding? What might these clusters represent in the context of molecular properties? 2. Compute the trustworthiness score of the t-SNE embedding. How does this metric quantify the quality of the embedding? Use the following code to calculate the score: from sklearn . manifold import trustworthiness score = trustworthiness ( X_scaled , embedding , n_neighbors =5) print (f" Trustworthiness score : { score }") 3. Discuss the challenges and best practices for tuning t-SNE parameters like perplexity and learning rate. 4. Why might t-SNE be particularly useful in chemistry applications? Provide examples, such as cluster- ing compounds or analyzing molecular properties. Submission Instructions Submit a report containing: • Python code for all parts of the problem. • Plots and visualizations. • Written answers to all questions and reflections on t-SNE’s performance. Hints: • Install scikit-learn if needed: pip install scikit-learn • Trustworthiness provides a quantitative measure of embedding quality. Problem 4: Understanding and Implementing UMAP UMAP (Uniform Manifold Approximation and Projection) is a dimensionality reduction technique that is particularly effective for visualizing high-dimensional data in low-dimensional spaces. This problem will guide you through understanding, implementing, and analyzing UMAP using Python. (a) Introduction to UMAP 1. What is UMAP? Research and provide a brief explanation of what UMAP does and how it works. Include a discussion of the following: • The mathematical foundation of UMAP (manifold learning, topology, etc.). • The main parameters of UMAP (e.g., n neighbors, min dist) and their effects on the embedding. • A comparison of UMAP with other dimensionality reduction methods such as PCA and t-SNE. 2. Why is UMAP particularly well-suited for visualizing high-dimensional datasets? (b) Dataset Preparation Download a high-dimensional dataset of your choice for analysis. For example: • MNIST: Handwritten digit images (available in sklearn.datasets). • ESOL, ChEMBL, or QM9: Chemical datasets containing molecular features. • MOSES: Molecular datasets with SMILES strings. Use the following Python code snippet to load the MNIST dataset as an example: from sklearn . datasets import fetch_openml import pandas as pd # Load MNIST dataset mnist = fetch_openml (’mnist_784 ’, version =1) X = mnist . data y = mnist . target print (f" Shape of data : {X. shape }, Shape of labels : {y. shape }") Answer the following questions: 1. What is the dimensionality of the dataset? 2. How would you preprocess this dataset for UMAP? Perform any necessary preprocessing steps, such as scaling or normalization, and provide the Python code. (c) Implementing UMAP Use the umap-learn Python library to reduce the dimensionality of your dataset to 2 dimensions for visu- alization. Here is a code snippet to get started: import umap . umap_ as umap import matplotlib . pyplot as plt # Initialize and fit UMAP reducer = umap . UMAP ( n_neighbors =15 , min_dist =0.1 , random_state =42) X_embedded = reducer . fit_transform. ( X) # Plot the embedding plt . figure ( figsize =(8 , 6) ) plt . scatter ( X_embedded [: , 0] , X_embedded [: , 1] , c=y , cmap =’Spectral ’, s =5) plt . colorbar ( label =" Digit Label ") plt . title (" UMAP Projection of MNIST Dataset ") plt . xlabel (" UMAP Dimension 1") plt . ylabel (" UMAP Dimension 2") plt . show () Questions: 1. What do the parameters n neighbors and min dist control in the UMAP algorithm? Experiment with different values for these parameters and describe their effects on the embedding. 2. How does the UMAP embedding compare with the original high-dimensional data? (d) Analyzing UMAP Results After generating the 2D embedding, analyze the results: 1. Identify any clusters in the 2D projection. Do these clusters correspond to meaningful patterns in the original data (e.g., digit classes in MNIST or chemical properties in molecular datasets)? 2. Compute the pairwise distances between points in the original high-dimensional space and compare them with distances in the 2D embedding. What can you infer about UMAP’s ability to preserve local versus global structures? 3. For chemical datasets, relate the UMAP clusters to specific molecular properties such as polarity or molecular weight. Are there clear separations between different types of molecules? (e) Comparison with PCA and t-SNE Perform dimensionality reduction on the same dataset using PCA and t-SNE for comparison. Use the following code snippets: from sklearn . decomposition import PCA from sklearn . manifold import TSNE # PCA pca = PCA ( n_components =2) X_pca = pca . fit_transform. (X) # t- SNE tsne = TSNE ( n_components =2 , random_state =42) X_tsne = tsne . fit_transform. ( X) # Plot PCA plt . figure ( figsize =(8 , 6) ) plt . scatter ( X_pca [: , 0] , X_pca [: , 1] , c=y , cmap =’Spectral ’, s =5) plt . colorbar ( label =" Digit Label ") plt . title ("PCA Projection of MNIST Dataset ") plt . xlabel (" PCA Dimension 1") plt . ylabel (" PCA Dimension 2") plt . show () # Plot t- SNE plt . figure ( figsize =(8 , 6) ) plt . scatter ( X_tsne [: , 0] , X_tsne [: , 1] , c=y , cmap =’Spectral ’, s =5) plt . colorbar ( label =" Digit Label ") plt . title ("t- SNE Projection of MNIST Dataset ") plt . xlabel ("t- SNE Dimension 1") plt . ylabel ("t- SNE Dimension 2") plt . show () Questions: 1. How do the embeddings generated by PCA, t-SNE, and UMAP differ in terms of cluster separation and overall structure? 2. Which method appears to work best for this dataset, and why? Consider factors such as local and global structure preservation, computational efficiency, and interpretability. 3. Reflect on the strengths and limitations of UMAP compared to PCA and t-SNE. (f) Applications of UMAP in Chemistry Provide examples of how UMAP can be applied to chemical datasets. Possible applications include: 1. Visualizing chemical space to identify clusters of similar molecules. 2. Analyzing high-dimensional molecular features for drug discovery. 3. Reducing the dimensionality of quantum chemical datasets for machine learning models. Discuss how UMAP’s ability to preserve local structure can be beneficial in each of these scenarios. Submission: Submit a report containing: • Python code for each part of the problem. • Visualizations of the UMAP, PCA, and t-SNE embeddings. • Written answers to all questions and interpretations of the results. • An analysis of UMAP’s applications in chemistry.
COM6016: Cyber Threat Hunting and Digital Forensics Forensics Case Study Assessment , October 2024 Submission Deadline: 15:00 on Monday, 16th December 2024 This assignment is worth 60% of the module mark. This assignment is made up of four different parts. You are required to answer all the questions below. All answers must be supported with adequate academic references. The document should be formatted using 12 point font size. The maximum number of pages for this assignment should not exceed 12 pages. PART 1 [20%] Sarah, a long-time employee at Spark Toys, was recently passed over for a promotion, leading to a decline in her morale. Shortly after, a significant data breach occurred, compromising sensitive company and customer information. Due to performance issues and suspected misconduct, Sarah was suspended and is currently under internal investigation for sending offensive messages. Her company-issued laptop has been seized and a memory image acquired as part of the investigation. Recent news indicates that Sarah has resigned from Spark Toys and accepted a position at a direct competitor. It is suspected she might have been involved with the data breach. Using your knowledge of Digital Forensics and the Digital Forensics process, describe how you would approach this case. You should ensure to discuss relevant information that could be retrieved from the memory of the device showing evidence of how this might be retrieved. PART 2 [45%] Xaiver is a staff member of CR BioTech, a company based in London at the forefront of cutting edge treatments for the flu. Xaiver is suspected of stealing chemicals and customer data from CR BioTech. She has also recently become a person of interest in an ongoing INTERPOL case involving the international export and sale of counterfeit cat flu medication. The counterfeit medication has been known to cause ‘gingivitis’ (inflammation of the mouth) and ulcers within three weeks of completing the suggested doses. Yusuf, one of Xaiver’s suspected accomplices who is now in custody, has suggested that the duo have made over £600,000 in sales of the counterfeit drug to more than 12 countries this year. Xaiver has been arrested and two USB drives have been retrieved from her. The disk images of the USB drives have been made available to you - USB1.E01 and USB2.E01 (attached on blackboard and also provided to you on the forensics laptop). Assume you work for PRISM forensics, an organisation providing forensics, first respondents and incident response services to various regional Police units and INTERPOL. You are required to write a maximum of a 5 page forensics report explaining how you went about your investigation and highlighting potential pieces of evidence that suggest that Xaiver was or was not involved in selling and exporting counterfeit drugs. PART 3 [15%] BridgePay, is a digital escrow payments service based in the UK. Their core application consists of a web application and SQL database hosted on various Ubuntu 18 servers. From the web front-end, staff of BridgePay can access an administrator-only area where they can view transactions made by customers. The web-based front-end and the mobile app can also be accessed by customers (buyers and sellers) using a web browser. On the 3rd of June 2024, the company went through a security audit and it was identified that some of its applications are vulnerable to ● CWE-434: Unrestricted Upload of File with Dangerous Type ● CWE-78: Improper Neutralisation of Special Elements used in an OS Command ('OS Command Injection') ● CWE-918: Server-Side Request Forgery (SSRF) On the 19th of October 2024, at 3pm, the company received an email from a third party claiming to have accessed its IT network and downloaded its customer's data requesting for a payment in bitcoin within three days to avoid public release of the data. Assume, you work for BridgePay as an incident response and forensics analyst, explain how you would go about handling this incident to ensure digital evidence is captured,forensics integrity is maintained and the business operations suffer minimal impact. PART 4 [20%] Your colleague, an IT administrator, suspects there is some suspicious activity going on, you have been provided a network capture. Using your knowledge of cybersecurity and network forensics, you are required to analyse the PCAP file 2024_part_4.pcapng and suggest what you think might be going on in the network packet sequence.
Econ 323 Comprehensive Problem Set (CPS) [Optional] This problem set will be made up of three blocks. Block 2 and 3 will be filled in later in the semester. This block corresponds to material before MT1. The CPS is optional. If a student elects to do it, it may replace one of the three midterm exams. That is, the CPS can replace a low midterm score, or if a student has not taken one of the exams for some reason, it can fill in for the missing test. As described in the syllabus and in class, the CPS is intended to be incrementally more challenging than the “standard” practice problems or exam problems we do routinely. You can think of it as a “stretch” assignment. All questions will have a small novelty or two that the student needs to work out on their own. The problems should then seem harder than our “standard” work. Block 1: Answer 2 of the following 3 questions. 50 points per problem. 1. A Monopsony is the lesser-known sibling of a Monopoly. While a Monopoly is a market with only one seller of a good, a Monopsony is a market with only one buyer of a good. This leads to its own kind of market failure. [Some context if you are interested: Monopsony comes up most frequently in labor markets. As an interesting example, professional sports leagues have many characteristics of Monopsonies. The NFL is the pretty much the only elite league for professional football. If you want to play pro football at a high level, the NFL is the only entity that will buy your services. This gives the NFL market power over players, and probably limits the amount the league pays for player salaries, through features like “salary caps” and other restrictions. [As an aside, I speculate that, since European soccer has many elite professional leagues, this probably leads to higher player salaries in soccer. Kylian Mbappe, for example, recently had a contract that paid $121 million per year over two years (with signing bonus) for PSG. Cristiano Ronaldo supposedly is being paid $220 million to play in the Saudi pro league, though his particular high salary is probably due to other factors. Meanwhile, Dak Prescott in the NFL is the highest paid player at “only” $60 million per year,] A company/factory town would be another example of a Monopsony. If there is only one significant employer buying labor in a local area, that employer will have a unique type of market power over its employees. Fewer jobs with lower salaries will be the result in a Monopsony, as you will discover.] The analysis for market failure with a Monopsony is similar to that of a Monopoly, with one key difference. Your goal in this problem, if you attempt it, is to figure out how that works. If you study the notes from our class on Monopoly, and try to apply our method to the Monopsony case, that will hopefully provide some hints on how to analyze this situation. Now for the problem: A market for a good is characterized as a Monopsony. There are many sellers and a single buyer of the good, the “Monopsonist”. The Supply curve is: PS(Q)= 100+1.8Q The Marginal Benefit to the Monopsonist from buying the good is PB(Q)= 2000-1.4Q a.) What is the efficient quantity and price of the good? b.) What is the equilibrium quantity and price of the good? c.) What is the Marginal Benefit to the Monopsonist at the equilibrium? d.) What is the deadweight loss created by the Monopsonist? 2. Beekeeping for the purpose of making honey provides external benefits to farmers located near the beehives, as bees pollinate their plants and improve crop yields. Bees also produce negative externalities. Being stung by bees is unpleasant. The more bees in a certain area, the more likely it is that people will be stung. In addition to hurting, some people have alergic reactions to stings and thus it can clearly be costly. Suppose we have the following situation. Marginal Benefit to Honey Consumers from an additional beehive: MPB=5000-75Q Marginal External Benefit to neighboring farmers from an additional beehive: MEB=1000+.2Q Marginal Cost to Honey Producers of an additional beehive: MPC=500+5Q Marginal External Cost to neighbors from additional beehives: MEC=0+.1Q a.) Find the market equilibrium Quantity and Price of hives. b.) What is the MSB at the Market equilibrium? MSC? c.) What is the socially optimal (Efficient) level for this society? d.) What is the Deadweight loss at the Market equilibrium? e.) What is the tax/subsidy that will lead to a new Market equilibrium at the Optimal quantity? 3. Suppose the government wishes to reduce the amount of methane gas, a waste product created from refining petroleum. Suppose the marginal benefit to society from reducing (abating) methane from the current levels is constant at $6000 per gigaton. Abatement levels will be on the x-axis and Costs on the y-axis. Assume that there are four producers, named 1, 2,.3, 4. The marginal abatement costs are: Producer 1: MAC,1 =.3A1. Producer 2: MAC,2= .5A2. Producer 3: MAC,3= .6A3 Producer 4: MAC,4= (1/3)A4 The total amount that is ultimately abated is A=A1+A2 +A3+A4 a.) The Marginal Abatement Cost (MAC) for the whole market is the horizontal sum of the individual producers’ marginal abatement costs. What is the market-wide MAC as a function of A? [Note, the coefficient on producer 4 is 1/3, which will give a beautiful clean answer in the end. If you round 1/3 to .333 or something be aware that it will be slightly off, but you can probably guess what the number is suppose to be.] b.) Given the answer to 1, what is the socially optimal level of methane abatement? Prior to any abatement, Producer 1 produced 9000 gigatons of methane, Producer 2 produced 10000 gigatons, Producer 3 12000 gigatons, Producer 4 produced 11000 gigatons. The total methane is thus 42000 gigatons. c.) The government wishes to use permits to achieve the optimal abatement. Each permit will allow the owner to produce 1 gigaton of methane. How many permits must the government issue to achieve the optimal abatement level? d.) If the permits were not tradeable, how much must each firm abate, and what are the Marginal abatement costs for each firm when they reduce their pollution to the permitted level? e.) If the firms can buy/sell their permits with one another, how much abatement does each firm do in the equilibrium?
ENGLISH LANGUAGE STUDIES ESL 1700 HIGH-INTERMEDIATE NELP 1700-A34 FALL 2024 COURSE DESCRIPTION ESL 1700 is a student-centered high-intermediate ESL class, which has two goals. The first goal is to provide you with the opportunity to further develop your abilities to read and write at a college level, while improving your English grammar and vocabulary. Secondly, this class will give you the opportunity to develop your academic studying skills and critical thinking through readings by authors, art critics, and designers. REQUIRED MATERIALS • Line Color Form, by Jesse Day • A bilingual dictionary with reliable information on parts of speech, countability of nouns, transitive and intransitive verbs, irregular forms, pronunciation, and example sentences. LEARNING OUTCOMES By the successful completion of this course, students will be able to: 1. Identify and write with basic rhetorical structures, relevant grammar and vocabulary to express their ideas. 2. Generate specific support within paragraphs and essays. 3. Identify and correct sentence structure errors, such as run-ons, fragments, and comma splices. 4. Recognize plagiarism and incorrect citation of outside sources. 5. Correctly format Chicago-style. footnotes and bibliography. 6. Read non-technical material for main ideas with minimal help from dictionaries. 7. Identify an author’s point of view and purpose. 8. Participate in informal discussions in and out of the classroom as well as give regular presentations without reading material. 9. Actively participate as an audience member of a presentation. 10. Follow directions, and ask for clarification from an instructor about assignments. COURSE REQUIREMENTS Participation and attendance Homework Presentation Formal writing assignments Tests and quizzes FINAL GRADE CALCULATION • Participation 40% of grade • Projects (papers 60% of grade and presentations) Tests, quizzes, in-class writing Weekly homework If you have too many latenesses, partial absences, and/or late or missed homework, I will lower your grade one step (for example, from an A- to a B+). COURSE OUTLINE Chronology and topics are likely to change according to the needs of the class. Additional assignments will be given on a week to week basis to align with ongoing work. Holiday: November 29 Thanksgiving Break DATES Topic H.W. Assignments WEEK 1 AUG 30 Introduction Writing assignment Writing: Personal essay WEEK 2 SEP 6 Vocabulary Grammar review – parts of speech Reading and grammar assignments WEEK 3 SEP 13 Academic writing overview The writing process - finishing Reading and grammar assignments WEEK 4 SEP 20 QUIZ 1 The writing process - editing Reading (Line Color Form) WEEK 5 SEP 27 Essay organization/thesis statements Reading strategies Writing: Description WEEK 6 OCT 4 QUIZ 2 Summary outlines Reading document 1 WEEK 7 OCT 11 Descriptive paragraphs Grammar topic Reading and grammar assignments WEEK 8 OCT 18 MIDTERM TEST Snapshots Writing: Opinion WEEK 9 OCT 25 Consultations Reading WEEK 10 NOV 1 Summary and paraphrasing Writing: Summary and response WEEK 11 NOV 8 QUIZ 3 Summary and paraphrasing Reading WEEK 12 NOV 15 Research: Summary, paraphrase and quotation Writing: Research WEEK 13 NOV 22 QUIZ 4 Research: Citations - basic CS rules TBD WEEK 14 DEC 6 Recurrent grammar and style issues TBD WEEK 15 DEC 13 FINAL TEST Recurrent grammar and style issues
MRI Project: T1 mapping in the heart 7MBI2003: Imaging in Healthcare 2024/25 1) Literature review Perform a literature search about T1 mapping in the heart, exploring clinical applications, challenges, and the main sequences used in this context such as MOLLI and SASHA. Please mention the differences between these sequences and standard T1 mapping sequences that you have seen in the MRI lectures. A good starting point for the literature review is to look at these articles: -Kellman P, Hansen MS. T1-mapping in the heart: accuracy and precision. Journal of cardiovascular magnetic resonance. 2014 Dec;16(1):2. -Taylor AJ, Salerno M, Dharmakumar R, Jerosch-Herold M. T1 Mapping: Basic Techniques and Clinical Applications. JACC Cardiovasc Imaging. 2016 Jan;9(1):67-81 You can also have a look at the provided slides for a brief explanation of cardic T1 mapping. The literature review should be part of your final written report. [20 marks] 2) Implementation of cardiac T1 mapping - In the programming language of your choice, implement cardiac T1 mapping functions that would allow you to calculate T1 maps from MOLLI data. - In the programming language of your choice, implement cardiac T1 mapping functions that would allow you to calculate T1 maps from SASHA data. [30 marks] 3) Results Run your implementation on the provided SASHA and MOLLI cardiac Dicom data and evaluate the obtained T1 values for each sequence by using regions of interest in the myocardium (heart muscle). Hint: in Matlab, you can use the “dicomread” and “dicominfo” functions and inversion times can be obtained for each Dicom image using info.TriggerTime where info is the struct obtained from the call of “dicominfo” . [30 marks] 4) Discussion Interpret the results you obtained (do the results agree with what is published in the literature? how do the accuracy and precision of these methods compare?) [20 marks]
ECON 400: MACROECONOMICS FOR MONEY, BANKING AND FINANCE - 2024-2025 COURSE AIMS & OBJECTIVES, KEY SKILLS, AND LEARNING OUTCOMES The course introduces state-of-the-art methods used in current macroeconomic research to understand short-run business cycle and inflation dynamics, as well as economic stabilisation policies. We will develop a broad and deep knowledge of modern Dynamic Stochastic General Equilibrium (DSGE) macroeconomic models that employ microeconomic foundations and rational expectations. These models will be solved using advanced analytical and numerical-computational approaches. More specifically, we will use the DSGE neoclassical Real Business Cycle (RBC) and New Keynesian (NK) frameworks to understand the different sources of aggregate economic fluctuations, and to examine the effects and roles of fiscal and monetary policies. Finally, the course examines contemporary issues such as financial frictions, government debt financing through distortionary taxation, and liquidity traps. Upon successful completion of this module, students should be able to: Analyse problems by identifying appropriate economic models and select the appropriate techniques to solve those problems. Understand the importance of micro-foundations and rational expectations in macroeconomic analysis. Understand the role of central banks and governments in periods of economic uncertainty. Use programming techniques to simulate and analyse Dynamic Stochastic General Equilibrium (DSGE) models. Seewww.dynare.orgfor the programing software that runs on MATLAB and that is used to simulate DSGE models. Use rigorous economic arguments to justify policy conclusions. You should strike a balance between reading the main textbooks and the journal articles listed below, and critically evaluate how these fit into the body of knowledge on the subject. Performing exercises, completing the coursework, attending (watching) lectures and workshops, revising, and working consistently throughout the year are vital for succeeding in this module (and any module as a matter of fact). ECON 400 provides a solid foundation for a macroeconomics dissertation, further postgraduate studies in economics, and advanced research projects in academia, economic think-tanks, policy institutions, and central banks. COURSE STRUCTURE There are overall 20 hours of lectures and overall 10 hours of workshops / lab sessions throughout the Michaelmas (First) Term. We will also have online Q&A sessions. See Online Timetabling for more information. FINAL MARKING INFORMATION This course is assessed by means of formal examination and coursework. The final mark for the module is calculated as: 33% Coursework + 67% Final Exam. COURSEWORK AND FINAL EXAM The coursework assessment for this module comprises of one compulsory assignment that will involve applying both the theoretical and computational methods taught throughout the lectures, workshops and labs. There is a non-binding word limit of 2,000 words. Note that while it is important to try and follow the word count, we will not penalize your work if you are under or over the limit (within reason of course). The assignment will be evaluated based on its' quality and not on its' length. Therefore, focus on answering the questions properly by using the technical tools learnt in the lectures, workshops and labs, and by employing solid economic intuition and references to support your answers. The assignment will be uploaded on Moodle by the end of Week 5 while the submission deadline is set for Week 10. The final exam normally takes place in January. The exam assessment consists of overall 3 questions. Students are required to answer any 2 out of the 3 questions. All questions are equally weighted. You should not submit answers to more than the required number of questions. If you do, we will mark the questions in the order that they appear, up to the required number of questions in each section. Please note that previous exam papers may not have operated under the same exam rubric or assessment weightings as those for the current academic year. The content of past papers may also be different. FEEDBACK ON COURSEWORK AND FINAL EXAM The coursework and final exams will be marked and returned to students with feedback comments within 4 weeks of the coursework submission deadline / final exam date. MARKING CRITERIA AND PENALTIES PLEASE NOTE: Work submitted up to three days late with no agreed extension will receive a penalty of 10%. Work submitted more than three days late will receive a mark of 0 (zero). The definition of ‘days late’ includes weekend days; deadlines will normally be set such that the third day does not occur at a weekend. If the third day falls on a weekend, students will have until 10.00 am on Monday to hand in without receiving a further penalty. Extensions will be given only in the most exceptional of circumstances. Marking criteria can be found in the MBF Postgraduate Handbook on the MBF programme Moodle page.
AS.440.601.58.FA24 Microeconomic Theory Your Signature Assignment will consist of a presentation of the analysis of a data set due week 8 class and will count for 13 percent of your final grade. In-class presentations will be conducted during weeks 9 and 14. The data set should have at least 100 data points and 3 independent variables. Good data sources are the data archives of various universities, government agencies, international financial institutions, etc. In all cases the source of the data must be provided. In preparing for your PowerPoint presentation, you are expected to do outside research or reading. However, the analysis should not replicate an existing published analysis. The empirical analysis should be done in MS Excel. Below is a suggested outline of your presentation: 1. The topic of your presentation 2. Importance of this topic 3. Key questions/issues you plan to address 4. Sources of information/data The presentation should be at least 10-12 slides total in length that describes the data, the question(s) the analysis has addressed and answered, the methods, and results. This can include regression equations, tables of regression coefficients, and selected graphs. The key is selection: this material should support your answers, not be a historical description of how you got to the final model(s). One guide is to visualize writing a concise memo to your general manager/CEO explaining what you found in enough detail to support your results and give her/him confidence that you have covered potential problems. In the appendix you may describe in more detail what you did. This should include any transformations made to the variables and why, construction of indicators or new variables, if any. The objective is to give more information on how you approached the data and how you structured your analysis.
HISTORICAL ESSAY MAT390/HPS390 FALL TERM 2024 The following description outlines the three course components related to your historical essay: DRAFT ESSAY: (5% of your total grade for this class, 3/3) A draft of your paper should be uploaded electronically to QUERCUS. Your draft essay will be graded based on the following (1 point for references/bibliography, 1 point for identifiable essay structure such as thesis statement introduction body and conclusion, 1 point for completeness in terms of length of the paper). If you don’tupload your draft essay to Quercus by the due date you will receive 0/3 on this component. No late submissions. You also must upload your draft a second time to Peer Scholar. To do this, click on the Peer Scholar exercise on our course page, load Peer Scholar in a new browser window, goto the Create phase of the activity and upload your draft document. PEER SCHOLAR EXERCISE: (5% of your total grade for this class, 2/2) Using Peer Scholar you will be matched with two essays written by your peers in this class. You will be given the task of reading these essays and offering your feedback on them. You have already practiced applying criteria to evaluate the quality of historical arguments in A2: Human and AI Generated Texts on History of Mathematics. You will apply similar skills here to essays composed by your current peers. The students whose work you comment on will then receive your feedback so they can revise their argument into something stronger. You must upload your drafts to Peer Scholar to participate and get marks for this activity!! NOTE: While your peers will help you identify areas to revise before final submission, student feedback is independent of grades given to you by course staff on your draft and final essay submissions. The grade you are given for this activity is based on the quality of your written and evaluative feedback. You have to complete all aspects of the “create” “assess” and “reflect” phases in Peer Scholar to get all points for the Peer Scholar activity. If you don’t finish all three phases you won’t get full marks. We will devote the activity period on Monday November 18 to completing the "assess" phase of the Peer scholar assignment. FINAL ESSAY: (25% of your total grade for this class, 25/25) Uploaded electronically to QUERCUS. The essay should be 2000 words (papers accepted in the range of 1900-2200 words). LATE POLICY: There will be a deduction of 5% per day (including weekend days) on later papers. This final essay is a major component of your grade for this course. It is recommended that a topic be chosen early and that your research and writing be underway by later October to early November. The essay should consist of your own work; it will be run through a standard database to verify that there has been no plagiarism from any source. In particular, do not cut and paste text from websites or Wikipedia articles into your essay. Do not plagiarize. POLICY ON USE OF GENERATIVE AI: This assignment is designed to be completed without the use of generative AI, using only concepts and skills we have developed in activities and through discussion of our course readings. However, as stated in the course syllabus, you may choose to use generative AI tools like ChatGPT as a tool when working on this assignment. We encourage you to review the section on generative AI tools in the course syllabus before considering using such tools within this assignment. Excellent work can be produced without the use of AI. The discussion we will have while completing A2: Human and AI Generated Texts on History of Mathematics will help us determine as a class, some ways generative AI can be useful and some of its pitfalls. Remember you are responsible for the final work you submit for this assignment. If you choose to use generative AI tools while working on this assignment, you must acknowledge which tools you used and how you used them. It is an academic offence to not credit sources—including generative AI—in any work you submit for academic credit. This acknowledgment should take the form of an appendix and in-text citations. See below for details. APPENDIX: Students who choose to use AI must submit an appendix with their historical essay. This appendix should be titled “Appendix: Statement on Usage of Generative AI” . Underneath this title, you must provide links to the raw transcripts that record your interactions with the tool while working on the assignment, including prompts and responses. For example, you should create a subtitle in your appendix such as “Compose an opinion editorial about how great Newton's discovery of calculus is” prompt. ChatGPT 3.5, 25 Sept. version, OpenAI, 3 Oct. 2023, chat.openai.com/chat. Then paste the text that was generated on that date by this prompt, followed with the link to the transcript. For ChatGPT transcripts,use the “share” icon in the upper right hand corner. You can find out more how to create these citations herehttps://style.mla.org/citing-generative-ai/. CITATION OF Appendix CONTENT PRODUCED BY GENERATIVE AI: Moreover, should content be produced, written or discovered through artificial intelligence and this research or text incorporated into the body text of the historical essay, students must cite that content. For all standard references in this assignment, Iamasking for those to be formatted in APA style. For your references that use information gathered through AI prompts, you will reference that part of your Appendix that has compiled all your prompts and responses generated. If you quote or paraphrase a passage from text generated by AI in response to a prompt, you should have already pasted the prompt and its associated text into your Appendix. Cite that output, signalling you arrivedata certain factor opinion or through your interaction with AI. You can do that by citing the Appendix using intext citation,e.g., place the citation (See Appendix: Statement on Usage of Generative AI, p. 13) at the end of the sentence of text in which you used those ideas gather from AI. Your final paper will be evaluated out of 25 points according to the following grading rubric: 7.5 points Motivation/Arguments: The essay presents a concise, well-stated, interesting and non-trivial thesis that is argued for persuasively. Analysis of historical sources is demonstrated. Clear and deep understanding of topic and concepts investigated. Student expresses reasoned opinion about topic at hand. Topic of paper is within scope of the course. 5 points Structure : The paper has an introduction, body, conclusion, and well-written topic sentences and coherent transitions between paragraphs. Historical material is presented in a logically cohesive way. Conclusion follows from the thesis and supporting evidence. Word count within 1900-2200 range. 5 points Written Style: Sentence structure to the point, appropriate use of paragraphs and foot/endnotes, formal (academic) style in mostly the third person, capitalization of proper nouns [e.g.: names, places, book titles], use of italics, underline or boldface is appropriate. 2.5 points Sources : At minimum the essay demonstrates a thorough reading and analysis of at least eight different sources. At least six of these sources must be academic publications that can be found through the search function of the University of Toronto library system. Application of APA bibliographic style is clear and consistent for all citations. Quotations from sources are appropriate. Any usage of generative AI as a discovery tool within essay construction has been properly cited and documented within an Appendix according to the “ POLICY ON USE OF GENERATIVE AI” laid out in this document. 5 points Overall effort : Essay demonstrates original thinking, creative research skills and/or good overall effort. The essay led the student to consider new questions or state a novel perspective on their topic. A passionately argued essay using mathematical and/or historical evidence. American Psychological Association (APA) bibliographic and citation conventions Please apply APA style. for citations and references: https://apastyle.apa.org/style-grammar-guidelines/references/examples Where to get started? The MacTutor website for the history of mathematics can be a starting place. If you goto the bottom of an article of interest, click >References >show to see the source list for links. This is a good way to start finding sources for your research, depending on your topic. It can be a useful place (and a more rigorous/academic online source than Wikipedia) for history of mathematics:http://mathshistory.st- andrews.ac.uk/ Essay Topics Your essay should develop a perspective upon (thesis) about your topic. General descriptive overviews are not useful. Evidence presented in the essay must always have relevance to your thesis claim. I also encourage you to express yourself in your writing. Connect yourself in some way with your topic – this makes your writing more meaningful. What interests you about the topic? Use this question as your “way in” to finding a claim you want to make about it. What surprised you during your research? Observations you make, that are unique to you, often lead you to the best thesis statements. Give us your take on this discovery. Although your essay will contain factual material, it should be focused, analytical and motivated by your argument about a particular view upon the given topic. Is history just facts? No! It is interpretation of the facts that makes some historical essays exceptional! HERE ARE THE TOPICS ON QUERCUS ARE LINKS TO RESOURCES THAT MAY HELP YOU GET STARTED ON EACH TOPIC AREA 1. Getting into their heads: What Babylonian mathematical artifacts tellus 2. How Greek and Roman people told time: Ancient Sundials 3. Are non-constructive proofs convincing? Archimedes’ proof of circular area 4. Does the Ancient Greeks practice of geometrical construction prove the existence of mathematical objects in Euclid’s Elements? 5. Not a Computer: The Antikythera mechanism and ancient Greek astronomy 6. How is the study of cosmic triangles in Ptolemy’s Almagest and Aryabhata’s Aryabhatiya, different from, or similar to, techniques of modern trigonometry? 7. In what ways was Japan’s wasan period of mathematics culturally distinct in style, practice, and topics investigated as compared with other mathematical traditions? 8. Transmission of knowledge or independent discovery? Pascal’s Treatise on the Triangle and Zhu Shijie’s Precious Mirror of the Four Elements 9. Differences and similarities between Francois Viète’s The Analytic Art and modern algebra 10. Simplifying calculation prior to the modern digital computer: Napier's invention of Logarithms 11. Logic of the highly improbable: How did Compte de Buffon, Blaise Pascal and Pierre de Fermat justify the quantification of chance? 12. Why was Cavalieri accused of heresy for his method of indivisibles? Did Cavalieri's method present a logically sound way of measuring areas and volumes? 13. Is mathematics invented or discovered? Solution to the Pell equation in India, China and Europe 14. "Recognize the lion by his claw": The role of competition in generating solutions to the Brachistochrone problem 15. A precursor to Tartaglia and Cardano? Expanding concepts of algebra in the Islamic golden age: the case of Omar Kayam and the cubic equation 16. A clock that keeps perfect time: Christian Huygens and the isochrone curve
CIVIL 507 - 2024 Term Project Controlling Structural Vibrations. Purpose. The purpose of this project is to investigate problems related to vibrations in structures. The nonlinear response of selected structures may also be investigated. A list of topics that can be investigated is provided below: 1) Vibrations induced by people: Pedestrian bridges; Floors with walking people; Floors for sport or dance activities; Floors with fixed seating and spectator galleries; High-diving platforms 2) Machinery-induced vibrations: Machine foundations and supports; Bell towers; Structure-borne sound; Ground-transmitted vibrations 3) Wind-induced vibrations: Buildings; Towers; Chimneys and Masts; Guyed Masts; Pylons; Suspension and Cable-Stayed Bridges; Cantilevered Roofs 4) Vibrations induced by traffic and construction activity: Roads; Railways; Bridges; Construction Work 5) Earthquake-induced vibrations: Buildings; Bridges; Dams; Towers; Chimneys and Masts; Underground structures. If there is a topic of interest to you that is not included in the list above, please discuss it with the course instructor. Project Requirements. A) Topic Selection: Select one structural system that maybe sensitive to vibrations for your investigation. A good starting point are the papers from the Journal of Sound and Vibration, the Journal of Earthquake Engineering and Structural Dynamics and the proceedings of the International Modal Analysis Conference (IMAC) and the International Operational Analysis Conference (IOMAC). Consult with the course instructor before proceeding with your term paper. B) Literature Review: Conduct a critical review of the literature regarding the vibration problems associated with the structure selected, and how such problems can be minimized or removed completely. Identify proposed methods to control or minimize vibrations and discuss their merits and limitations. C) Case Studies: Select two (2) case studies to illustrate how a vibration can be minimized or controlled. You may use simple hand calculations or more sophisticated computer analysis to show how your method is effective in controlling vibrations. Provide a clear description of each problem and discuss the general applicability of the solution method used. All assumptions must be justifiable and clearly stated. You can also select a proposed method of vibration control and demonstrate why the method does not work, or the conditions under which the method may work. D) Report: Your project report should be prepared individually, although it is encouraged to discuss your findings and solutions with your classmates. Your report will be submitted in electronic form. only, and will consist of two parts. The two parts of the report should be electronically submitted via CANVAS. E-mailing the report to the course instructor is permitted, but if the files are too big they should be submitted in a CD or USB memory stick. The first part is your report about the literature search and a summary of your findings. The two case studies will also be part of this report. If computer analyses are conducted as part of this work, the input and output files should also be submitted. The suggested length of the body of your report is not more than 25 pages. Copies of papers and reference material can also be submitted as appendices. There is no limit on the number of pages included in the appendices. The second part of your report will consist of a PowerPoint presentation that summarizes in 12 slides or less the work that you have done. Submission deadline. The submission date of the report is 6th December at 23:00. Reports submitted after the deadline will be deducted 10 marks/day Project Evaluation The submitted report will be evaluated in terms of its presentation and content. Marks will be assigned as follows: Structural Dynamics Term Project - Evaluation Sheet Student Name: Presentation (15%): Organization (3%): % Readability & Clarity (5%): % Composition (spelling & grammar) (4%): % Quality of graphics (3%): % Contents (85%): Introduction (10%): % Literature Review (15%) % Report contents (35%): % Objectives and Scope Statement of Problem Mathematical Approach Discussion Illustrative Example (20%) % Originality Clarity Practical relevance Summary / Discussion (5%): % Other (Supporting Info.) (10% extra) % Tardiness (reduction: 10 marks per day): % TOTAL: %
CSE 240 Homework 08, Fall 2024 (50 points) Due Saturday, December 7th, 2024 at 11:59PM, plus one day grace period Introduction The aim of this assignment is to make sure that you understand and are familiar with the concepts covered in the lectures on Prolog. By the end of the assignment, you should have • strong concept of logic / declarative paradigm; • strong concept of facts, rules, and questions in Prolog. • Compared and contrasted the programming C++, Scheme, and Prolog. • Prolog list operations and manipulations • Prolog recursive rules with multiple nested calls Reading: Text Chapter 5 and lecture slides. Also read Text Appendix B.4 Prolog Tutorial You are expected to do most of the assignment outside the class meetings. Should you need assistance, or have questions about the assignment, please contact the instructor or the TA during their office hours. You are encouraged to ask and answer questions on the course discussion board and in the course Discord server. (However, do not share your answers in the course discussion board.) Practice Exercises (no submission required) 1. Answer the multiple-choice questions in text section 5.9 that have been covered in the lectures. 2. Complete the questions 2 through 5 in text section 5.9. 3. Lookup the Unix command table in Appendix B.1 and read the Prolog tutorial in B.4. 4. Log onto general.asu.edu and execute the Unix commands and Prolog exercises. 5. Follow the Prolog tutorial in Text Appendix B.4 or the tutorials given in the course Web site, start GNU Prolog and try following simple programs (build-in functions). Don't forget the period at the end of the statement. | ?- write(hello) . % hello is a constant */ | ?- write(Hello) . % Hello is avariable. Its address will be displayed*/ | ?- write('helloworld') . % a string is printed */ | ?- read(Y), write('The variable entered is '), write(Y), nl. /* nl prints a newline. Type a period and an enter at the end of the input */ | ?- X is 2+2. | ?- Y is 5*8. | ?- Y is 2**10. | ?- length([a,b, x, y, 2, 45, z], L). | ?- append([a,b, c, d], [4, 6, 8], LL). | ?- append(X,Y,[a,b,c]). Then, type ";" to obtain all possible answers. | ?- X is [1 | [2 | [3 | []]]], write(X). Explain the output. 6. Use trace and notrace before and after your questions in Question 5. For example | ?- trace. | ?- X is [1 | [2 | [3 | []]]], write(X). Explain the output. | ?- notrace. Programming Assignment (50 points) 1. Consider the following diagram of a familytree: Implement the family relationships in a Prolog factbase: Note: The section in the highlighted box above has been completed for you. /* Factbase for a family tree. It consists of facts and rules. */ /* Facts */ male(abe). male(rob). male(jim). female(joy). female(ana). father_of(abe, ana). /* abe is the father of ana */ father_of(abe, rob). /* abe is the father of rob */ father_of(abe, jim). /* abe is the father of jim */ mother_of(joy, rob). /* joy is the mother of rob */ mother_of(joy, jim). /* joy is the mother of jim */ mother_of(joy, ana). /* joy is the mother of ana */ /* Complete the facts given in the diagram above */ /* Rules */ Open the given file 1FamilyTreeSolution.pl using a text editor such as pico or vim on a Unix operating system. You may write the program on your own computer using any editor such as notepad and upload the program to the general server to compile and execute. Compile the program using the command: $ gplc 1FamilyTreeSolution.pl This should create an executable that you can then run using the following command: $ ./1FamilyTreeSolution This will enter you into the GNU Prolog programming environment where your program has already been executed. Ask questions by typing, e.g.: |?- father_of(mark, beth). |?- mother_of(beth, tom). To exit from GNU Prolog, type the end-of-file character at the main Prolog prompt ^d (Ctrl-d). |?- ^d You can find a complete set of GNU Prolog commands at: http://www.gprolog.org/manual/gprolog.html For all the following questions, please label the question number in a comment. For example, if Question 2.1 asks you to define a rule named is_male (X) that returns "yes" (or "true") if X is the father of a member of the family, then your code should look like: /* Question 1.1 */ is_male(X) :- male(X); father_of(X, _). This is a reading check. If you see this, congratulations. You may use this in your solution. Now, you can start to add your code into the program. Complete the program by adding facts for the remaining members on the familytree. The section inside of the highlighted box has already been completed for you for clarification. Please pay close attention when adding the remaining family members. spelling counts and all letters should be lowercase. 1.1 Define (add into the factbase) a rule called is_male(X) that returns “yes” (or “true”) if X is a male or the father of a member of the family. Note: the system will return a "yes", if it finds a "true" answer and there exist no more true answers. The system will return "true ?" if it finds a "true" answer and there are still possibly further matches. In this case, if you press enter, it will return "yes" and stop. If you type " ;", it will continue to search for further answers. 1.2 Define (add into the factbase) a rule called is_female(X) that returns "yes" (or "true") if X is a female or the mother of a member of the family. 1.3 Define a rule called grandmother_of(X, Z) that returns “yes” (or “true”) if X is a grandmother of Z. Define another rule called grandfather_of(X, Z) that returns “yes” (or “true”) if X is a grandfather of Z. (you can use 1.5 parent_of(X, Y) rule) 1.4 Define a rule called sibling_of(X, Y) that returns “yes” (or “true”) if X is a sibling of Y. Note: a family member cannot be their own sibling. (you can use 1.5 parent_of(X, Y) rule) 1.5 Define a rule called parent_of(X, Y) that returns “yes” (or “true”) if X is aparent of Y. 1.6 Define a rule called descendent_of(X, Y) that returns “yes” (or “true”) if X is a descendent of Y. A descendant, in the context of a family tree, refers to any individual directly related through a series of parent-child relationships, e.g. if joy is the parent of jim, and jim is the parent of kim, then kim is considered a descendant of both joy and jim. Note: you will need to use recursion as well as the parent_of rule. (you can use 1.5 parent_of(X, Y) rule) 2. Consider the following factbase for the games planetDrop and mechaTech. [12 points] planetDrop mechaTech genre: action genre: rpg focus: focus: gameplay story - mechanics - characters - framerate - plot graphics world - rendering - design - meshes - culture In 2GamesSolution.pl, complete the following: Create facts for the genres: genre(X, Y) where X is a game and Y is a genre. Create facts for the focuses: focus(X, Y) where X is a game and Y is a focus. Create facts for the elements: element(X, Y) where X is a focus and Y is an element. 2.1 Create a rule game(X, Y) where X is a game and Y are the genre and the focuses of that game. For example, game(planetDrop, X) should return (use ; to browse through the results): X = action ? ; X = gameplay ? ; X = graphics 2.2 Create a rule creation(X, Y) where X is a game and Y are the elements of that game. For example, creation(planetDrop, X) should return: X = mechanics ? ; X = framerate ? ; X = rendering ? ; X = meshes 3. In 3QuicksortSolution.pl, you will re-implement the Quicksort given in textbook and lecture slides. In the given example, the first (left-most) element of the given list is selected as the pivot. In this question, you must choose the second element of the list as the pivot. Hint: You can represent the input list into pairs: [First|[Pivot|Tail]]. You must write comments to indicate the size-n problem, stopping condition and its return value, size m-problems, and construction of the size-n problem from size-m problems. Test case: | ?- qsort2([8, 3, 4, 12, 25, 4, 6, 1, 9, 22, 6], Sorted). It returns: Sorted = [1,3,4,4,6,6,8,9,12,22,25] 4. A premium pizza is comprised of exactly 40 ounces of toppings. The available toppings are listed below with their corresponding weight (in ounces). There can be multiple entries of each topping, as long as the total weight is equal to 40 ounces. Topping Weight (ounces) Pepperoni Sausage Bacon Onion Mushroom 4 10 6 5 7 For example, a pizza can contain 1 topping ofpepperoni, 2 toppings ofsausage, 1 topping of bacon, and 2 toppings of onion : 1*4 + 2*10 + 1*6 + 2*5 = 40 (ounces) A pizza cannot contain 7 toppings of bacon : 7 * 6 = 42 > 40 A pizza cannot contain only 3 toppings ofsausage : 3*10 = 30 < 40 In 4PizzaSolution.pl, define a rule pizza(P, S, B, O, M) to find out how many of each topping can be contained on a pizza, where arguments P, S, B, O, and M represent the weights (in ounces) of the Pepperoni, Sausage, Bacon, Onion, and Mushroom toppings, respectively. Your pizza rule will be used to find the outputs to questions similar to below: Test Case: | ?- pizza(1, S, 1, O, M). It returns: (use ; to browse through the results) M = 0 M = 0 M = 0 M = 0 O = 6 O = 4 O = 2 O = 0 S = 0 S = 1 S = 2 S = 3
SOC 20004/30004: Introduction to Quantitative Methods in Social Science Final Project Due December 12, 2024 This assignment is based on the Early Childhood Longitudinal Study Kindergarten Cohort, 1998 (ECLSK 98), a nationally representative sample of US children entering kindergarten in 1998. The assignment follows from a preliminary analysis you completed for assignment 2. In that study, you looked at bivariate relationships between the social origins of a child and that child’s early mathematics skill. We chose this outcome because it is known to predict later school success – which itself predicts adult economic outcomes. What is new in this final project is the task of modeling these relationships. The data set name on canvas is “ECLS_98_Soc30004_revised.dta.” We’ll confine our interest to the following variables: B_MOMEDW = mother’s education POV= family’spoverty status ln_income (treat as interval scale) R4MTHT_R=child’s math score in fall of kindergarten R4MTHT_0=child’s math score in spring of kindergarten Please write a report not to exceed 10 double-spaced pages (including tables and figures) that answers the following questions: 1. To what extent do social origins (mothers ’ education, family poverty status, and income (log scale) predict a child’s math scale when the child enters kindergarten (typically at age 5)? 2. To what extent does the child’s math skill at entry to kindergarten predict that child’s math skill in the spring of kindergarten, controlling for the social origin variables? 3. All findings in social science are based on some assumptions regardless of how carefully the analysis is done. Please state any assumptions that are required to justify your answers to questions (1) and (2). Also, please evaluate the credibility of these assumptions with respect your findings in this study. (Use plots to evaluate these assumptions when possible.) Your report should define key variables used in the analysis and provide a table of basic descriptive statistics. Please use multiple regression to answer these questions. Please write down any models that you study. Define each term in the model. Please describe key findings with language accessible to a policy maker or practitioner. (Although pvalues are relevant, do not rely primarily on computed p values to interpret your findings. If a relationship is “statistically significant,” interpret the magnitude of the regression coefficients themselves.) In a final paragraph, briefly summarize your key conclusions.
2024-2025 ASSESSMENTS Undergraduate Workbook 2: Responsive and Accessible Web Design SCC.306 Internet Applications Engineering Recommended Completion Time [30 Hours] Assessment Weighting [20%] Learning Outcomes Details of the Learning Outcomes of this module can be found in the Module Programmes catalogue: https://portal.lancaster.ac.uk/intranet/mpc/modules/016886/000123. This tables specifies which learning outcomes are assessed and in which way they are assessed. Subject Specific Learning Outcomes: Knowledge, Understanding and Skills Outcome Covered How is it covered Understand web architectures, standards, and business practices. Yes Students need to understand the web standards that exist, how to explore what they standardise and then apply that knowledge. Empirically assess the performance of a web site No Understand and alleviate potential performance bottlenecks. No Address issues and limitations of scale. No Accessibility and Internationalisation Yes Students need to develop websites that are accessible and broadly usable. Design for Heterogeneous platforms (Responsive web design) Yes Students need to develop a website that is response. Establish a quality process for web sites Yes Students need to understand how to ensure the quality of a website. Security threats and hardening of web sites No General Learning Outcomes: Knowledge, Understanding and Skills Outcome Covered How is it covered Establish performance metrics No Interpret quantitative data to identify performance problems No Make informed choices about complex distributed and networked architectures No Feedback and Deadline Feedback on this workbook will be provided continuously in the labs. Please ensure you attend them. The feedback on your submitted workbook will be a response to the entire cohort. If you wish to receive more detailed individual feedback, you will need to go through the cohort-wide feedback and identify and comment on which pieces of feedback do/do not apply to your work. You can then contact Phil and Matthew to discuss your comments on the feedback. The expected return of feedback is 4 weeks after the coursework deadline. The deadline for this work is Friday 4pm Week 9. Introduction In this workbook, you will be looking at both responsive and accessible web design. In Part 1, you will focus on how to build an accessible website, whilst Part 2 looks at how to approach responsive web design. It is important that you read all instructions given in this coursework specification. For both parts of the exercise, you will use and build upon the same template website. In other words, once you have completed Part 1, the same website will be used for Part 2 – there is no need to start again. The template website is provided on Moodle. The template consists of these files: · index.html · style.css · adysonbeneath-the-screen-638106667407073410.jpg · the-lancaster-history-and-creative-writing-lecture-2024-638651199449566120.jpg · location-map.svg · lu-logo.svg · pint-of-science-morecambe-637848439194681824.jpg · qs-top-150.svg Submission Once you are finished with both parts of this workbook, please submit all of your code and this workbook (as a single zipped folder) using the submission point provided in Moodle. You should not need to submit a copy of the website for both Part 1 and Part 2. A single copy of the website after Part 2 should be sufficient. If during Part 2 you implement additional features that changes the layout away from the specified layout, then you should include a second copy of the website that contains those additional features. References This is an academic piece of work and you are expected to use correct academic citations. No specific style. is mandated; however, you should be consistent and use the same style. throughout this document. The library provides a guide on the Harvard style. (https://lancaster.libguides.com/harvard) and resources are available for SCC-specific tools (https://lancaster.libguides.com/computing/referencingtools). If you use (or adapt from) code taken from third party sources (i.e., code that you have not written and code that we have not provided to you), then you must cite where this code has come from (use the same referencing style. as for this document). You should include references in your source files and explain how code has been adapted from its source. Part 1: Accessible Web Design The first part of this workbook is to examine the accessibility of the supplied website. This involves identifying the issues that are causing the web page to be poorly accessible to some audiences, and then fixing a subset of these by modifying the web page. To begin with, we recommend you review the W3C web content accessibility guidelines https://www.w3.org/TR/WCAG22/ and the BBC’s Standards and Guidelines for Mobile Accessibility: http://www.bbc.co.uk/accessibility/forproducts/. You may also find the following resources useful: · http://wave.webaim.org/ · https://www.w3.org/WAI/standards-guidelines/ · https://www.w3.org/WAI/standards-guidelines/wcag/ · https://www.w3.org/WAI/standards-guidelines/wcag/faq/#start · https://www.w3.org/WAI/fundamentals/accessibility-intro/#examples · https://www.w3.org/WAI/fundamentals/accessibility-principles/ · https://www.w3.org/TR/WCAG22/ · https://developer.mozilla.org/en-US/docs/Web/Accessibility Throughout this section you should refer to appropriate accessibility standards or guidelines by citing them. You may choose which citation style. you use, but should be consistent throughout the workbook. You must use academic reference styles, simply providing hyperlinks is insufficient. An example entry in your bibliography might look like: 1. 1.4.3 Contrast (Minimum) - Level AA (2019) Web Content Accessibility Guidelines 2. W3C Web Accessibility Initiative. Available at: https://www.w3.org/WAI/WCAG21/quickref/#contrast-minimum (Accessed: October 1, 2023). You can then cite this in-text like this (WCAG2 1.4.3). Typically, you would cite this as (WCAG2 2019a) and then other references from WCAG2 as (WCAG2 2019b) etc. However, we are happy for you to use the standard number in the in-text citation. Alternatively, you could number each reference and cite it like so [1].
Module 8 Final Paper: Health Intervention Overview The final paper culminates the reading assignments, lectures, and application activities. It will allow you to apply the concepts, facts, and information you learned throughout the course. As described in the instructions below, you will develop an intervention to address social determinants of health in ONE marginalized community of interest. The proposed intervention should address levels of the socio-ecological model. You should refer to the application exercise worksheets in previous modules (M3 - M7) to help generate ideas for the final paper and ensure you meet the requirements for this assignment. You should think of this final assignment as an opportunity to create an intervention to increase health equity in a community that experiences health disparities at multiple levels (e.g., communities and healthcare systems). This is the time to dream and plan for what is possible for those in need of health equity. Instructions Choose ONE community from one of the application exercises you previously completed. You will expand this into a paper in which you provide a detailed intervention. Write a 1,200-1,500-word, double-spaced paper using social determinants of health and thesocio-ecological model to create an intervention to increase health equity among a marginalized community (e.g., racial and ethnic minorities, older adults, people with disabilities). Provide an overall description of the marginalized community who will benefit from the intervention. In your paper, make sure to incorporate the completed table from the application exercise that corresponds to the community you are focusing on. The purpose of incorporating the table is to provide a visual aid and concise overview of the social determinants of health and levels of the SEM. Make sure the table describes the specific social determinants of health that affect health outcomes for the community you've chosen and that you plan to address with the proposed intervention. You can screenshot the table and add it to your paper. You should then elaborate on the intervention after incorporating the table into your paper. Your intervention must include more than one level from the socio-ecological model (e.g., intrapersonal and community). Your intervention must also explicitly state which social determinants of health you are addressing at each level from the socio-ecological model. Your paper must also describe the intervention and how it addresses the stated social determinants of health within each of the identified levels of the socio-ecological model. The intervention must also include cultural competence or cultural humility in the design.
Department of Mechanical and Industrial Engineering MIE1620: Linear Programming and Network Flows (Fall 2012) Quiz 2 December 5, 2013 1. (17 points) Consider the uncapacitated network flow problem shown in Figure 1. The label next to each arc (in boxes) is its cost. The bold numbers beside the arrows are the supplies/demands. Consider the spanning tree indicated by the dashed arcs in the figure and the associated basic solution. (a) (4 points) Determine the values of the arc flows corresponding to this basic solu- tion. Is the solution feasible? Nondegenerate? (b) (3 points) For the basic solution from part (a), find the reduced cost of each arc in the network. (c) (4 points) Determine an optimal solution to the network flow problem. (d) (3 points) By how much can we decrease c45 and still have the same optimal basic feasible solution? (e) (3 points) If we increase the supply at node 1 and the demand at node 5 by a small positive amount δ, what is the change in the optimal cost? Figure 1: Network for Problem 1 2. (10 points) Consider two nonempty polyhedra P = {x ∈ ℜn | Ax ≤ b} and Q = {x ∈ ℜn | Dx ≤ d}. We are interested in finding out whether the two polyhedra have a point in common. (a) (2 points) Formulate an LP problem with the following properties: if P ∩ Q is nonempty, the LP returns a point in P ∩ Q; if P ∩ Q is empty, the LP is infeasible. (b) (3 points) Form the dual of the LP from part (a). (c) (5 points) Suppose that P ∩ Q empty. Use the dual from part (b) to show that there exists a vector c such that c′x < c′ y for all x ∈ P and y ∈ Q. 3. (6 points; 3 points each) Consider a linear programming problem of the form. (a) Form the dual of the problem. (b) Explain how Dantzig-Wolfe decomposition can be applied to the dual. It is suffi- cient to provide a few sentences to identify the coupling constraint and to describe the constraints of the subproblems solved during a typical iteration is sufficient. 4. (9 points) Consider a transportation problem with two source nodes s1 , s2 , and n demand nodes 1, . . . , n. All arcs of the form (si , j), i = 1, 2; j = 1, . . . , n are assumed to be present and to have infinite capacity. Let D =Σ di be the total demand. Let the supply at each source node be equal to D/2. Assume that di > 0 for all i. (a) (3 points) Draw a figure representing the network flow problem. Include all nodes, arcs, supply and demand arrows, etc. (b) (2 points) How many basic variables are there in a basic feasible solution for this network? (c) (4 points) Show that there exists a degenerate basic feasible solution if and only if there exists some set S ⊂ {1, . . . , n} such that Σi∈S di = D/2. 5. (8 points) Consider the following LP Supposex(ˆ) is a feasible solution to formulation (1). Now consider an inverse formulation to formulation (1), which we assume to be feasible: We wish to understand the structure of optimal solutions to formulation (2). (a) (2 points) Prove that formulation (2) has an optimal solution. (b) (3 points) Determine if p = 0 can be a part of any feasible solution (c, p, ϵ) to formulation (2). (c) (3 points) If at an optimal solution (c* , p* , ϵ* ), the constraint A(x(ˆ) − ϵ* e) ≥ b is satisfied, show that there must exist some i such that ai(′)(x(ˆ) − ϵ* e) − bi = 0.
Civil 507 - Dynamics of Structures I 2024 Course Outline Acknowledgment Statement: UBC’s Point Grey Campus is located on the traditional, ancestral, and unceded territory of the xwməθkwəy̓ əm (Musqueam) people. The land it is situated on has always been a place of learning for the Musqueam people, who for millennia have passed on their culture, history, and traditions from one generation to the next on this site. Dynamic Analysis and Response of Structural Systems: • formulation of equation of motion • free vibrations and response to harmonic, periodic and transient loadings • response to impulsive loadings and response to general dynamic loadings • numerical evaluation of dynamic response • earthquake response of linear and nonlinear systems • generalized SDOF systems • frequency domain method of analysis • formulation of equations of motion for MODF systems • free vibration of MODF systems • dynamic analysis and response of MODF systems • damping in structures Recommended Texts: • A.K. Chopra, Dynamics of Structures: theory and applications to earthquake engineering, Prentice Hall Inc., Englewood Cliffs, New Jersey, 5th Edition 2016. • R.W. Clough and J. Penzien, Dynamics of Structures, Computers and Structures Inc., Berkeley, California, 3rd. Edition, 2003. • J.L. Humar, Dynamics of Structures, CRC Press/Balkena, The Netherlands, Third Edition, 2012. Suggested Computer Programs: Maple, Mathcad or MATLAB, NONLIN, SEISMOSIGNAL, SAP2000 Course Schedule: The course lectures will be delivered through one lecture per week on Monday between 2:00 to 5:00 p.m. A detailed schedule of the course, including deliverable deadlines will be posted on Canvas at the start of the course. Course Mark: 30% mid-term take-home exam, 40% final take-home exam and 30% for a term paper. The mid- term exam will be held during the first week of November, and the final take-home exam will be held during the exam period at the end of the term. The term paper is due the last day of classes. The projects will be based on chapters 1 to 12 of Chopra’s book. The practice problems shown on the last page do not need to be handed-in, but can be submitted to the TA for review and marking. Solutions to some of the assigned problems will be discussed in class.
Debt Financing and Management for Public Organizations Final Project Fall 2024 For this project you will answer questions about an actual municipal issuer, the New York State Thruway Authority (the “Authority”), and a Request for Proposals (“RFP”) issued by the Authority in 2021. At the time, the Authority was seeking banks to serve as underwriters on future bond transactions, as outlined in Section 1 B of the RFP (“Purpose”) and Section 2 (“Engagement Requirements”). They are asking proposing banks to demonstrate two main qualifications: 1. Credentials of the proposing bank that highlight recent experience underwriting similar municipal bonds including impressive rankings, investment banking staff as well as sales/trading capabilities and capitalization of the bank; and 2. Knowledge/deep understanding of the Authority’s structure, debt management strategies and credit strengths. If you were working at an investment bank that is interested in winning underwriting business with the Authority you would need to outperform. all other competing firms on both of the contests above. When reading the RFP, you will note that there are extremely strict requirements for the response materials, formatting and even the font and paper size utilized. Any deviation from these requirements will disqualify a firm from being awarded the underwriting business. Section 4A describes how banks will be selected. The criteria against which proposals will be evaluated is also listed in Section 4C. Note that the scoring rubric for RFP responses is in part tied to New York State procurement rules (because the Authority is a State agency) including an advantage for underwriting firms that are a certified Minority- or Women-Owned Business Enterprise. To set your firm apart in the stack of proposals, your submission must excel on objective #2 above – proving that you have a thorough understanding of the Authority’s current operations, the structure and relative strength of their individual bond credits, their outstanding debt and future financing needs. To study the Authority you should start by reading the most recent Official Statement for Authority bonds and all of the continuing disclosure documents posted publicly to the MSRB’s EMMA website. At the time, this RFP was released the most recent transaction was the Series O bonds priced in September 2021 via competitive sale. Locate the OS for the Authority Series O bond transaction (CUSIP# 650009W58). As a study guide to help prepare the best possible underwriting proposal, answer the following questions. Note: You may perform. a google or news search on the Issuer to help describe incurred or expected challenges for this organization but keep in mind that any events which occurred after September 2021 would not have been known or relevant in 2021. Focus on press, rating reports (posted to the Final Project Assignment on Brightspace), and annual continuing disclosure documents posted to EMMA in 2021 or before. For the calculations discussed in Question 15, please use the Excel file posted to the Final Project Assignment on Brightspace containing information on the Authority’s outstanding debt, including the Series O bonds. This is the same data printed in the Series O Official Statement, transferred to Excel so you do not need to type in the figures. Answers to Questions 1-15 must be submitted in an unlocked copy of this Word document, INCLUDING YOUR FULL NAME AT THE TOP. FINAL PROJECT QUESTIONS A Study of NY STATE THRUWAY AUTHORITY – The Issuer 1) Is the Authority a conduit borrower or a municipal entity that is able to issue tax-exempt debt directly? 2) Select text from the Official Statement that outlines the mission of the organization, summarized in no more than one paragraph. Include quantitative measurements about the Authority’s assets and operations. 3) There are many different liens that the Authority utilizes for financing of roads and bridges – see RFP Section 1A for a description of each. What type of pledge did the Authority provide for its Series O bonds? a) Is this a General Obligation or Revenue Bond credit? b) Are the Series O bonds issued under the most senior lien or a junior (subordinated) lien? c) What is the Security and Source of Repayment of the Series O bonds? Hint – use the “Revenue” section from “Source of Payment & Security” in the OS [pdf pg 16] 4) There is a Debt Service Reserve Fund provided for the Series O Bondholders (see OS Section “Senior Debt Service Reserve Fund” under “Sources of Payment & Security”). What is the Debt Service Reserve Fund Requirement, in terms of the minimum amount of money that must be maintained in the DSRF? 5) The Series O bonds have a Rate Covenant and an Additional Bonds Test. (a) Summarize the rate covenant that includes a Debt Service Coverage Ratio. Consider using the very short description provided in the Moody’s Rating Report. (b) Summarize the terms of the Additional Bonds Test – again, consider using the Rating Agency’s description. 6) Now that we know the debt service coverage ratios required for compliance with the bond covenants, what is the current Debt Service Coverage Ratio (DSCR) at the time the RFP was released? Focusing on Calendar Year 2020 results, the most recent fully completed fiscal year at the time of the RFP/Series O Sale, use $ figures reported in the Official Statement “Results of Operations” [Series O Official Statement PDF pg 30]. a) What are the Authority’s Net Revenues for 2020 (in $ Million)? b) Bonus Point: What has been deducted from the Total Revenues to calculate Net Revenues? c) What is the 2020 Net General Revenue Bond Debt Service (in $ Million)? d) What is the General Revenue Bond Debt Service Coverage Ratio for 2020? e) How does the 2020 DSCR compare with the minimum ratio for compliance with the bond covenants? f) What does the Moody’s rating report say is a DSCR for senior bonds (like the Series O) that would put the Authority at risk for a downgrade? g) Bonus Point: The toll revenues in 2020 were much lower than prior years. Before you interview with the Authority for underwriter RFP you should research why revenues were trending unfavorably and then propose an approach for marketing future bonds to investors in the context of current underperformance – ditto with the approach for maintaining the credit ratings. Read the OS Section entitled “Management’s Discussion & Analysis of Results of Operations” and summarize the Thruway’s explanation of why revenues were down in 2020. h) Bonus Point: If you have been hired by the Authority as an underwriter, you will be selling new bonds and helping make a market for trades of the Series O bonds in the secondary market. It’s important to keep a close eye on the Authority’s financial performance. What is the DSCR for the Senior Bonds as of today and has performance improved? Hint: look at Continuing Disclosure documents on EMMA for Calendar Year 2023 performance (most recent report available at this time) 7) Now that we know what conditions and financial performance are required for the Authority to issue additional debt, what are the projected future financing needs? Look at the forecast for capital expenditures announced by the Authority at the time of the Series O bond offering; see the Official Statement Section entitled “2021-2025 Capital Program”. a) State the total amount of spending or “investments” planned over the period of 2021-2025 (in $ Billion). b) Bonus Point: State the amount of the capital program (in $ Billion) that will be funded with General Revenue Bond Proceeds (the same pledge as the Series O bonds). Hint: Look at the Official Statement Section entitled “Funding of the 2021-2025 Capital Program” 8) The 2021-2025 Capital Program covers three major projects. In order to submit a thoughtful response to the Authority’s RFP you should be aware of all ongoing projects at the Thruway and be able to communicate how these projects might relate to future financing needs, the best type of debt to align with each project’s expected economic life and the Authority’s sources of repayment for related debt. List (but do not describe) the 3 major projects under the 2021-2025 Capital Program. 9) If selected to serve as the Authority’s underwriter, you will be responsible for outlining the Continuing Disclosure Obligations the Thruway has committed to, as well as confirming that the Authority has been in compliance with their other outstanding Continuing Disclosure Obligations at the time you sell any new bonds. Read the OS section entitled “Continuing Disclosure Under SEC Rule 15c2-12”. a) What are the Authority’s Continuing Disclosure requirements under the Series O Bonds? b) Bonus Point: Has the Authority been late, incomplete or otherwise materially noncompliant on any Continuing Disclosure Postings? Hint: this is something that must be disclosed in the OS. 10) If you are competing to win the Authority’s underwriting business, you’ll need to demonstrate that you can convince investors to purchase the Thruway’s bonds, which will inevitably entail answering questions investors have about the risks facing the Authority. Read all of the “Investment Considerations” the Authority reports on in the Series O Official Statement. List at least 5 of the main categories of risks, concerns or external factors that impact the Authority’s business but do not go into detail on the status of each. a) Bonus Point: One of the Authority’s recognized Considerations for Investors stated in the Series O Official Statement is that the Thruway credits could be downgraded. Is the credit rating on the Series O bonds the same, downgraded or upgraded since the bonds were sold in 2021? Hint: Look at EMMA. b) Bonus Point: Another one of the Authority’s recognized Considerations for Investors stated in the Series O Official Statement is legislative interference. Can the State of New York prevent the Authority from setting or collecting fees or charges? Series O Bonds as a Proxy for Authority Preferences: Plan of Finance, Bond Structure & Amortization 11) Review the bond maturities on the Series O transaction. Assuming the Series O is typical of the bond structure the Authority is most comfortable selling to finance toll roads and bridge projects, what are the following features of the Series O transaction that you should mimic when proposing future (new money) bond transactions in your RFP response – answer the questions below based on the features observed in Series O: a) Fixed Rate or Variable Rate Bonds? b) Longest maturity (in years from the date of sale)? c) Premium, discount or par bonds? d) Bonus Point: what do you know about the Authority’s financed projects that make the bond features above a good match for the Thruway’s needs? 12) You are preparing the RFP response which requires competing investment bankers to suggest a plan of finance for future Authority debt issuance. Read all of the intended uses of the Series O bond proceeds to determine if the Authority typically uses tax-exempt bond proceeds to finance the following needs – answer yes or no based on the observed use of the Series O bond proceeds: a) Deposits to the Debt Service Reserve Fund b) Costs of Issuance c) Capitalized Interest 13) Similarly, look at the Series O bonds to determine what tax exemption was awarded to the General Revenue projects to make assumptions about whether or not future bonds sold under the same credit will also likely be granted tax exemption under each of the following jurisdictions - answer yes or no based on the tax status of the Series O bonds: a) Federally tax-exempt? b) State tax-exempt? c) New York City tax-exempt? 14) You need to start modeling bond cashflows for the Authority’s future debt in order to answer some of the RFP questions. Assume that you will use the same interest payment dates as the Series O bonds if future bonds are sold under the same lien. List the semi-annual interest payment dates of the Series O bonds. 15) Generate two graphical summaries of the Authority’s outstanding debt to understand typical amortization preferences at the Thruway: a) Stacked bar graph of annual debt service on Series O with shading to differentiate principal vs. interest payments. Note that the Series O annual debt service figures have been provided in the Excel file on the Final Project page of Brightspace (see columns C & D). Create a graph in the same style. as in the in-class exercise template (posted to Brightspace “Class Sessions” as “IN CLASS 12 Excel Exercises” on the tab called “Graph of Series DS”). Paste the stacked bar graph below. Make sure all legends are labeled as are the graph axes. b) Stacked bar graph of annual debt service on Series O Bonds separate from the annual debt service on all of the Authority’s other Outstanding Debt with shading to differentiate Series O and the Other Outstanding Indebtedness. Note that you will need to combine the Other General Revenue Bond Debt Service from the Excel file Column B with the Outstanding Junior Indebtedness (Column G) to represent the Authority’s debt OTHER THAN SERIES O. Create a graph in the same style. as in the in-class exercise template (posted to Brightspace “Class Sessions” as “IN CLASS 12 Excel Exercises” on the tab called “Graph of Consolidated DS”). Paste the stacked bar graph below. Make sure all legends are labeled as are the graph axes. c) Bonus Point: based on the observed patterns in the Authority’s Series O debt service, which of the debt management strategies did the Thruway use when structuring the Series O bonds around other outstanding debt? Choose one of the styles we studied in class that most closely resembles the Series O debt service placement with respect to other outstanding debt: Uniform, Fill or Wrapped placement. d) Bonus Point: Do you have any recommendations for the Authority in your RFP response about a better approach to amortization or debt portfolio management, based on the Authority’s outstanding debt?
