Using Machine Learning to Predict the NBA’s Most Improved Player Award

This project explored the viability of machine learning techniques to predict the winner of the NBA’s most improved player award. This project had a heavy focus on preparing data for the modeling stage, however, this preparation was critical. Six models were trained and tested on data on the  1987-2023 seasons. The best models predicted correctly for 78% of the seasons, including the final 9. Viability for use as part of a sports betting strategy was also explored.