Please accept the [GitHub Classroom Assignment](https://classroom.github.com/a/gJ7smGl3).

The goal of this competition is to predict reviewer ratings of music releases (albums, EPs, etc.). You should make these predictions using the set of various features provided. This is an individual competition. AI tools such as ChatGPT are allowed.

You may use any classical method except neural networks / deep learning on this assignment - it is up to you to research whether a library makes use of deep learning (**please do not ask on discord if a given library is allowed**). Any methods that you use that were not covered in class must be fully cited and you must explain how you used all external and online resources.

Additionally, no boosting methods (including XGBoost, LGBM etc.) or libraries are allowed.

In addition to submitting your fully reproducible solution to your github repo, your repo should also contain a README.md explaining what you did for each of the data science steps:

1.    Data Exploration
2.   Feature Extraction / Engineering
3.   Model Creation and Assumptions
4.   Model Tuning
5.   Model Evaluation / Performance
6.   Struggles / Issues / Open Questions

This competition ends at 4:30pm EST on Monday, Nov 3rd. The private leaderboards will be made available during class where we will announce winners and prizes. You can finish up your README after class until 11:59pm when you must submit it to gradescope. **Please include your kaggle username in the README**

This competition uses **Macro Averaged F1-Scoring** as the evaluation metric. You only get 5 submissions per day because it's expected you do your own local evaluation instead of overfitting to the kaggle leaderboard. This cannot be changed - make sure to start early.

The top 2 public submissions will be used for your final score. If you want to select a different two submissions (that aren't your best public scores) you can but I don't recommend it. If you overwrite this default, you understand that these are the submissions that will be used instead - no changes to these can be made after the competition ends.

Grading will be based on the benchmark you hit, and the closeness you achieve towards the next benchmark. Points may be awarded or deducted based on the state of the repo, the steps / process you followed to get to the benchmark, and the reproducibility of one of your best scores (which should be the only code in your repo once the competition ends).

Note: Gradescope does not automatically sync with GitHub. If you submitted your repo to gradescope and made changes to the repo these will not be reflected in gradescope. You must re-submit. Expect this submission process to take some time so please make sure you give yourself at least an hour before the deadline.

**Note: DO NOT GIT ADD / COMMIT CSVs or DATA TO GITHUB.**