Applied Machine Learning
scikit-learn, keras, and tensorflow.
All proprietary information has been removed to protect organizational confidentiality.
This repository showcases shareable projects displaying advanced machine learning techniques in finance and operations.
1 Forecasting Expected Financial Returns
Practicum Project, College of Engineering at the University of Oklahoma
This practicum applied machine learning to forecast a key economic indicator for investment strategy. Using LSTM, Elastic Net, and Prophet, LSTM delivered the strongest results, cutting out-of-sample error by ~10%. The workflow used vintage data, time-based validation, and feature engineering to reflect real-world constraints.
2 Predicting Operational Delivery Times
College of Engineering at the University of Oklahoma
This project built models to improve food delivery time estimates by integrating traffic, accidents, and weather with historical data. The approach reduced variance in arrival predictions versus baseline methods, creating a more reliable framework to support delivery services with better planning accuracy and operational consistency.
View on GitHub:
