For this project I took a data set of over 8,000 mushroom samples and used 22 different features to see if the model could predict if the sample was either poisonous or edible.

I was able to work with both Linear Regression and XGBoost to see how well the model could predict each mushroom’s edibility. But more than that I was able to look at both permutation importances and shapley plots to see how the model made its predictions.

You can see the full explanation here on Medium.