Predicting Flight Delays: Logistic Regression
Developed a binary logistic regression model to classify U.S. domestic flight arrival delays using over 1 million records from January–February 2024. Independently sourced and cleaned the dataset, engineered features, built and evaluated the model using confusion matrices and ROC-AUC in R, and produced a written analysis for a non-technical professional audience. This project demonstrates end-to-end analytical ownership: from raw data to actionable conclusion.