What it does

The web application which uses machine learning techniques to pick the winner of an nba game. After testing 5 different algorithms, linear regression predicted the matches at the highest accuracy(65%)

Challenges I ran into

One of the largest challenges was consolidating all the data in an easy to manage form. NBA team data is difficult to find and is missing values. After cleaning the data the next step was to attempt to predict it by trying multiple machine learning algorithms. The last challenge was combining the work of the prediction with the front-end web application to create an interactive tool.

What's next for Hoop Labs

Hoop Labs would like to improve the prediction accuracy in the future and include more basketball data visualization.

Built With

Share this project: