Category: Uncategorized

Be sure your work is labeled by task number. Be sure to subm…

Be sure your work is labeled by task number. Be sure to submit your notebook file before the exam times out. All questions are worth 3 points, except as noted. Three point are awarded for appropriate use of markdown and documentation. Tasks 1. (5 pts) Read in the TechSales.csv file into a data frame called TechSales.  TechSales.csv TechSales contains records on sales representatives from the hardware and software product groups of a high-tech company. For each employee, the data include socio-demographic and education information, salary, sales performance, and a personality indicator. Also included in the data is a net promoter score, which is an indicator of customer satisfaction with each sales rep. This is a simplified representation of a realistic data set. The variables: •    Rep: A unique ID for each sales representative.•    Business: One of the two product groups: Hardware and Software•    Age: Employees actual age•    Gender: Male or Female•    Years: The number of years the employee has been employed at the company•    College: Whether or not the employee has a four-year college degree.•    Personality: One of four different personality types•    Certificates: The number of relevant professional certifications each employee has earned.•    Feedback: The average feedback score that each employee receives from his or her peers and supervisor on the 360 degree annual evaluation.•    Salary: Annual base salary of each employee•    NPS: The net promoter score (NPS) is a key indicator of customer satisfaction and loyalty. Explore the data. In particular, display the first few rows, determine the number of rows and columns, the data types of the columns and anything else you might want to/should know. There are no missing values or erroneous duplicates, so you don’t need to check for these.You do not need to rename the columns to adhere to Python standards. 2. Change column data types as appropriate. 3. How many employees are in each personality type? You should be able to output these values using a single command.  4.  How many employees of each personality type are in each business? You should be able to output these values using a single command.  5. Is a women more likely than a man to be in software as opposed to hardware?  You should be able to output the values that let you answer this question using a single command. Be sure to explain how you arrived at your answer. 6. Are gender and business statistically independent? How do you know? Provide your answer and support for your answer in a markdown cell. Support for your answer should be two or three probabilities and some verbiage. I don’t want a long story here. 7. Below is the correlation matrix for the relevant variables in Tech Sales. And here is a partial output of the statistical regression of Salary vs. all the predictor variables. You can see that Years has pretty much no correlation with Salary, yet the variable is (quite) significant in the regression model. Related to this is that the confidence intervals for the coefficients are rather narrow. Provide the most likely reason for this to be the situation. 8. (2 pts) Economic globalization is defined as the integration of national economies into the international economy through trade, foreign direct investment, capital flows, migration, and the spread of technology. Although globalization is viewed favorably by many, it also increases the vulnerability of a country to economic conditions of other countries. An economist predicts a 55% chance that country A will perform poorly and a 30% chance that country B will perform poorly. There is also an 18% chance that both countries will perform poorly. What is the probability that country A performs poorly given that country B performs poorly? Show your work and provide your answer to two decimal places. 9. (2 pts) Following up on the previous question, what’s the probability that country B performs poorly given that country A performs poorly? Provide your answer to two decimal places. 10. (2 pts) Do your answers to the previous two questions tend to support the notion that recent internationalization has made countries more vulnerable to economic conditions of other countries? Briefly support your response. JHouseDataThreeTownsS26.csv 11. Read in the JHouseDataThreeTownsS26.csv file into a data frame called housing. The file contains data describing over1,000 homes sold in three university towns. The meaning of each of the variables should be clear. Your objective will be to use regression modeling to predict the sale amount.. Begin by doing the usual data exploration. Also, describe the data. You don’t have to say much. Note that again there are no missing values or duplicate data. 12. Create the appropriate heat map of the correlations. (Try to omit the variable ‘record’.) 13. Create the scatterplots between the numeric variables between the numeric variables. You should be able to do this with a single command. Again, try and omit the column ‘record’. 14. Evaluate, without doing the regression, whether town is likely to be a useful predictor of sale_amount. 15. With an eye towards creating a regression models, what do you conclude/what observations can you make after exploring the data and viewing the charts produced in the last two tasks? 16. Which of the three values for town would be preferable to be the base value in the regression. Why? 17. Create a statistical regression, predicting sale amount from the other variables. You should use your answer to the previous question to specify the base value for town. (If you can’t do this, don’t spend a lot of time here.) 18. Provide a brief evaluation addressing:•    The statistical significance of the dependent variables.•    The possibility of multicollinearity. In particular are the signs of any of the independent variables “going the wrong way?” •    You do not need to address the assumptions of regression here beyond the above. 19. In this model, all other things being equal, how much does the predicted sale_amount increase for one additional square foot of home area?  20. In this model, on average how much more (or less) do homes in Eugene cost, compared to comparable homes in Lincoln. 21. With backwards stepwise regression in mind, suggest what your next step might be. In particular what model, if any is reasonable to try next and why? 22. Using the statistical backward stepwise approach, develop your best (final) regression model. Be sure to show your work. Provide an explanation as to how you arrived at this model and why it’s your solution to the problem. If you feel like this initial model is the best one then indicate that, and provide support for this decision. 23. (6 pts) How well or how poorly does the model follow the assumptions of regression? (It’s ok if not all the assumptions of regression are well-satisfied.) Provide support for your answer.  24. How good is your model? How good is your model in terms of prediction? 25. (5 pts) Now do a machine learning regression, using the best model you just arrived at above. You should use a test size of 20% and a random_state value of 77. 26. Evaluate the performance of your model. That is, how good is it in terms of prediction? 27. Provide a plot of predicted vs. actual values, with the regression line through the data. Summarize what you see in a sentence or two. 28. Provide the coefficients of the model. Is this model consistent with the statistics-based regression? (I’m not looking for anything quantitative here for this question – just an overall sense of things.) 29. (4 pts) Perform 10-fold cross validation linear regression using your regression model.  30. (5 pts) Evaluate the summary measure results both in the context of your previous results and independent of them.