Axel Perez

Date of Award

9-18-2021

Document Type

Thesis - SCU Access Only

Publisher

Santa Clara : Santa Clara University, 2021.

Degree Name

Master of Science (MS)

Department

Computer Science and Engineering

First Advisor

Yuhong Liu

Abstract

The amount of data that is being collected in all facets of life has increased astronomically. This trend has not escaped the world of sports, specifically in the National Basketball Association (NBA). Every team in the NBA now collects data on not only the players and teams in the league, but also players around the world as they look to scout the next up and coming NBA stars. Statistical analysis has taken over the league and has now become a prerequisite for almost all team and league staff positions. The NBA and many of its teams have invested in the infrastructure needed to collect large quantities of data. The quantity of data that is now collected can be daunting to work with and can be easily misused or can inquire significant computational cost. In this thesis, research was conducted on two dimensionality reduction techniques: Principal Component Analysis (PCA) and Autoencoder Neural Networks, which can capture the most important information in a lower dimensionality latent space, to then be used for downstream applications. High information containing embeddings are then constructed that are used to predict the net rating of 5 player lineups. Through the research and experiments of this thesis, the potential value of using deep learning techniques is demonstrated when creating player representations and making predictions with many features.

Download

SCU Access Only

To access this paper, please log into or create an account in Scholar Commons using your scu.edu email address.

Share

COinS