Improving PER (Player Efficiency Rating) in Basketball through Machine Learning

View Polygence scholar page
Project description
This project explores the intersection of advanced statistical methodologies and basketball with a focus on improving the Player Efficiency Rating (PER) metric. This research delves into three distinct AI models: Lasso Regression, Random Forest Regression, and Neural Networks. These models, each with unique capabilities, allow for more accurate PER ratings which helps teams and coaches to make informed decisions about player rotations and substitutions.

Project outcome
Presentation for Symposium of Rising Scholars and Research Paper for Journal of Sports Analytics, Curieux Academic Journal, and Research Archive of Rising Scholars

Mentor review
My mentor was extremely knowledgeable in her field and was a great listener who helped and gave me thoughtful advice throughout the entire research paper process.