juicob / predicting-nba-wins Goto Github PK

This project was created in to explore the NBA data set and create a machine learning classification model to predict the outcome of NBA games.

• Built With
• Python
• Plotly
• Pandas
• Numpy
• Dash
• Power BI

I wanted to see who had the most wins out of the entire dataset, so I grouped up all the teams and split them up between each conference and saw that my home team of the San Antonio Spurs came out on top followed by the Miami Heat!

Each team did significantly better playing at home versus playing at away games.

I've created two interactive dashboards one with Plotly/Dash (link coming soon!) and another with Power BI shown below. There's a lot to play with so have fun and let me know what you think!!

I created a Gradient Boosting Classifier and reached 82% accuracy based completely on features like assists, blocks, steals, and turnovers.