NBA Game Score Prediction with XGBoost This repository contains Python code for an NBA game score prediction model using XGBoost and a Flask application to display the predictions. The program uses nba_api to get data and fixtures from Stats.NBA

Functionality Data Acquisition and Processing:

Retrieves a list of NBA teams.

Queries the live NBA scoreboard for games.

Fetches historical game logs for a specified season.

Cleans and prepares the data for modeling.

Feature Engineering and Model Training:

Extracts features from game logs relevant to predicting points scored.

Trains an XGBoost regression model to predict game scores.

Evaluates the model's performance using various metrics.

Prediction and Flask App:

Loads the trained XGBoost model. Predicts points for upcoming games based on their team's historical stats. Provides a Flask application to display the predicted scores in a user-friendly format.

How to Use

Prerequisites:

Python 3.x with libraries like pandas, numpy, XGBoost, scikit-learn, requests, tabulate, and Flask installed. nba_api Steps:

Clone the repository. Run the script: python nba_game_prediction.py This will print the predicted scores for today's games on the console.

Web App: Navigate to the project directory in your terminal. Run the Flask app: python nba_game_prediction.py --run-app Access the predictions at http://127.0.0.1:5000/ in your web browser. Note: Depending on your environment, you may need to adjust the Flask app execution command.

Dependencies nba_api pandas numpy XGBoost scikit-learn requests tabulate Flask File Structure