NBA Game Prediction Model
2023
Predictive models for NBA game outcomes, reaching 66% accuracy — trained, evaluated, and compared Logistic Regression, SVM, Random Forest, Gradient Boosting, Ridge Classifier, and Neural Network models.
- Python
- scikit-learn
A model-comparison project on NBA game outcomes: Logistic Regression, SVM, Random Forest, Gradient Boosting, Ridge Classifier, and a Neural Network were trained and evaluated against each other on historical game data, reaching up to 66% prediction accuracy.
The focus was less on squeezing out a single best model and more on understanding how different model families trade off bias, variance, and interpretability on the same dataset.