top of page

Bookmakers hate them! Predicting outcomes of European soccer matches using linear regression

Reid Buzby & Michael O'Herron

Introduction

The goal of this project was to develop a model that could predict the outcomes of European soccer matches based on the characteristics of the teams involved in each match. We used a stochastic gradient descent linear regression algorithm to predict the number of goals scored by each team, which informed a prediction which team would win. We also used an SVM to classify matches into home team wins, losses, and draws directly. Lastly, we experimented with using an elementary machine learning pipeline by first using our regression model to predict the number of goals each team would score in the match, and then use those predictions as features to help the performance of the SVM classifier.

​

Click the buttons below to look at our process.

Here is a link to our codebase:

Here is a link to our full written report:

About
Contact
bottom of page