Machine Learning Projects
This repository showcases a series of machine learning assignments and projects that demonstrate my proficiency in various machine learning techniques, ranging from clustering to regression and neural networks. The projects are organized below, with the Final Project highlighted as the capstone of my learning journey.
- Objective: Analyzed salary fairness in Major League Baseball using machine learning techniques.
- Key Techniques:
- K-Means Clustering: Grouped players into performance archetypes (Power Hitters, Balanced Hitters, Utility Players).
- Neural Networks: Predicted salary fairness with high accuracy.
- Random Forest: Categorized players as underpaid, overpaid, or fairly paid.
- Insights:
- Identified salary inequities, with younger high-performing players often underpaid.
- Provided actionable strategies for teams to optimize payrolls while maintaining competitiveness.
- For more details, refer to the Final Project Document.
- Implemented and evaluated ensemble learning methods, including Bagging and Boosting.
- Demonstrated how ensemble models improve predictive accuracy compared to standalone algorithms.
- Applied Principal Component Analysis (PCA) to reduce dimensionality of complex datasets.
- Used K-Means clustering to identify patterns and group data points effectively.
- Explored Ridge and Lasso regression techniques to handle multicollinearity and perform feature selection.
- Compared model performance across different penalty parameters.
- Developed classification models using SVM to separate complex datasets.
- Tuned hyperparameters to optimize performance and evaluated results on test data.
- Analyzed time series data to forecast trends and patterns.
- Implemented ARIMA models and evaluated their accuracy in predicting future observations.