README.md

SwishPredict

A Data Pipeline for NBA Box Score stats from basketball-reference.com
Explore the docs »

View Demo · Report Bug · Request Feature

Table of Contents

1. About The Project
   • Built With
2. Getting Started
   • Prerequisites
   • Installation
3. Usage
4. Roadmap
5. Contributing
6. License
7. Contact
8. Acknowledgments

About The Project

1. Data Pipeline Architecture

• Uses Apache Airflow for orchestration
• Docker containers for modular services
• MinIO for object storage
• PostgreSQL for data persistence

2. Main Components

• Data Ingestion Service: Scrapes NBA player box scores from basketball-reference.com
• Data Processing Service: Processes raw data for model training
• Feature Generation Service: Prepares features for the ML model
• This pipeline is made to scrape box score data from basketball-reference for any NBA season with the month ranges October - July. Unusual seasons like 2019-2020 may call for adjustments to the scraping_config.yml config at data_pipeline_services/config/data_ingestion/scraping_config.yml

3. XGBoost (IN PROGRESS)

• Uses XGBoost for predicting fantasy basketball points
• Jupyter notebooks for model development and analysis stored in notebooks folder
• model is stored in models folder
• Model metrics and configuration are stored in config/model_metadata.yaml:

performance_metrics:
  test_mae: 6.163228800313608
  test_mse: 66.63370955961138
  test_rmse: 8.162947357395574
  test_r2: 0.7177536081562295

(back to top)

Built With

• 
• 
• 
• 
• 

(back to top)

System Architecture