Bayesian Estimation of Flexible Models for Sports Data

This project aims to explore flexible regression models for sports data. We aim to benchmark the two alternative approaches and study if it is more beneficial to have a single complex component (negative binomial) than to have several simple ones (Poisson). We will also explore finite mixture models as they offer a flexible alternative that may accommodate overdispersed data. The specific applications we will consider include, but are not limited to the following count data: number of goals, number of wins, and number of injuries. In each case, we are interested in predicting future outcomes, and also understand which factors drive the changes in the response variable. We will also consider classification problems, for example predicting performance of players and spotting talents. In particular, for the latter, understanding which factors contribute to future success is of great importance for team management.

Simran Bindra

University of Technology Sydney

Simran Bindra is a second-year student at the University of Technology Sydney. Having always had a strong interest in mathematics, he is pursuing a degree in data analytics while majoring in statistical modelling. Simran’s research project aims to combine his area of study, particularly Bayesian methods, with his other passion, sports, to accurately predict future outcomes of games and individual player performances.

You may be interested in

Joshua Rogers

Joshua Rogers

Accurate Multi-Gas Emissions Equivalence Modelling of Greenhouse Gases
Rebecca Rasmussen

Rebecca Rasmussen

What Kind of Random Walk are these Biological Cells Doing?
Alex Oakley

Alex Oakley

Appraisal of Machine Learning Evaluation Techniques in the Context of Glacier-Generated Events
Sean McGowan

Sean McGowan

PDE Discovery Using Machine Learning
Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text.