Basketball_ML

A machine learning project that combines a basketball game simulation with polynomial regression to predict optimal shooting parameters.

Overview

This project consists of three main components:

1. An interactive basketball game where players can manually shoot baskets
2. A data collection and model training system using polynomial regression
3. A testing environment to validate the trained model

Project Structure

Basketball_ML/
├── training_game.py      # Interactive basketball game for data collection
├── poly_regression.ipynb # Jupyter notebook for model training
├── test_model.py      # Game environment for testing the trained model
├── data.csv             # Collected game data
└── model.pkl            # Exported trained model

Features

Training Game (training_game.py)

• Interactive basketball shooting game
• Manual control of shooting parameters (speed and angle)
• Automatic data collection after each shot
• Data stored in CSV format with the following parameters:
  • Speed
  • Angle
  • Rim_center_x (basket position)

Model Training (poly_regression.ipynb)

• Jupyter notebook environment for data analysis and model training
• Uses scikit-learn library for polynomial regression
• Processes collected game data from data.csv
• Exports the trained model as a pickle file

Model Testing (test_model.py)

• Testing environment for the trained model
• Imports the trained polynomial regression model
• Allows validation of model predictions in the game environment

Getting Started

Prerequisites

pip install numpy pandas pickle pygame scikit-learn

Usage

1. Data Collection

   python training_game.py

   • Play the game manually
   • Data will be automatically saved to data.csv

2. Model Training

   • Open poly_regression.ipynb in Jupyter Notebook
   • Run all cells to train the model
   • The model will be saved as model.pkl
   • move the model to teams folder(create one if it doesn't exist)

3. Testing

   python Test_model.py

   • Test the trained model's predictions