This project aims to predict MLB team wins based on key statistics using machine learning models such as Linear Regression, Random Forest Regressor, and K-Means Clustering.
Dataset
Teams Dataset: Contains overall team statistics per season.
Batting Dataset: Aggregated individual batting statistics.
Pitching Dataset: Aggregated individual pitching statistics.
Fielding Dataset: Aggregated fielding statistics.
Approach
Data Preprocessing
Load and clean team and player statistics datasets.
Aggregate relevant statistics by team and year.
Identify key features that influence team wins.
Modeling
Linear Regression: Initial model to assess feature importance.
Random Forest Regressor: More robust model to capture non-linear relationships.
K-Means Clustering: Unsupervised learning to explore team groupings based on performance.
Results
Evaluated model accuracy using R² Score and RMSE.
Analyzed which statistics influence team wins the most