Collection of Assignments for CSE 3000: Contemporary Issues in Computer Science and Engineering
Note: Odd-numbered assignments are in HuskyCT, while even-numbered assignments are found in this repo and can be submitted to Gradescope via a Github connection.
Your goal is to build a machine learning model using given data and a Python template. You'll make changes to the GradientBoostingClassifer in mod02_build_bot_predictor.py and then test your changes in mod02_test_bot_predictor.ipynb. Can you minimize the misclassification rate on the test set?
You are given a model that was built to predict a mystery variable using zip code, income, and education level. The variable "group" represents a demographic that is prohibited and was NOT used in the model. Your goal is to use the SHAP package to review this model for bias.
Your goal is to connect anonymized data back to its source data and check the match rate. To complete this activity, write functions in the "deanonymize" script to (1) link the records in the two datasets and (2) calculate the rate at which you are successful doing so.
You will be given a toy dataset and have the opportunity to explore the potential effects of choosing sustainable energy sources to perform computations. You'll code the energy calculation in "sustainability calculations" script, and then you can review the results of various decisions in the "check calculations" notebook.
