The purpose of HW4 is to give you practice with the simple regression, multiple regression, classification, and resampling methods.
- README.me: this file
- ISSUE_TEMPLATE.md: grading rubric
- HW4.ipynb
- Default.xlsx: data to be used
More details are in the Jupyter notebook for HW4. Follow through the notebook and complete all your work there. When you are done, check back in your work. I will check out your completed work and grade it and return to you via check-in when your grade is complete (at the bottom of this notebook).
Note that I must be able to execute all cells in the notebook starting from top to bottom. I will not "clear all output" from your notebook and then execute it all in order to make sure the notebook works.
See cues in my skeleton notebook so you can get some information about what I am looking for on each section
| Part | Earned | Possible | Notes |
|---|---|---|---|---|
| 1. Short answer | | 1 | | |
| 2. Logistic regression| | 1 | | |
| 3. Simple linear regression | | 1 | |
| 4. Multiple regression | | 1 | |
| 5. Bootstrapping | | 1 | e.g., late penalty |
| Adjustments | | | e.g., late penalty,
messy code |
| Total grade | | | |
Like Dave explained - comments in python (using # in a code cell - or just adding a markdown cell) is critical so I can follow your work. You can't really over-comment things.
My expectation is that everything you type in a Python notebook is your own work. Any instances of "copy-paste" from the web or another person's notebook should be clearly cited. Of course you may look at examples but it is my strong preference that you refrain from copy-paste and type everything in. There is a learning reason for this, which I am happy to discuss in office hours or on slack.
I will follow UW academic misconduct policy for any suspected instances of cheating on HW or projects. Any confirmed instance of cheating results in a zero on a HW assignment. Any 2nd confirmed instance of cheating results in a zero for the entire course grade.