A machine learning project that predicts the NBA MVP (Most Valuable Player) based on player statistics from the last 20 years, achieving an accuracy of 81%.
This project uses machine learning to predict the NBA MVP based on historical player performance data. The model is trained on 20 years of player statistics, enabling it to identify trends and patterns associated with MVP-winning seasons. The primary goal is to leverage statistical data to make reliable MVP predictions with a high level of accuracy.
- Programming Language: Python
- Libraries: Pandas, Selenium, Beautiful Soup, Scikit-LearnProgramming Language: Python Libraries: Pandas, Selenium, Beautiful Soup, Scikit-Learn
- Automated Data Collection: Web scraping using Selenium to retrieve up-to-date player statistics from https://www.basketball-reference.com/.
- Data Parsing and Cleaning: Parsing HTML using Beautiful Soup and cleaning data with Pandas to convert raw data into a well-structured dataset.
- Model Training and Prediction: Implemented ridge regression and random forest models from Scikit-Learn to predict MVP outcomes.