NYC Traffic

Project Title: NYC Traffic & Trip Duration Predictive Model

Objective: Developed a machine learning pipeline to accurately predict transit times and traffic flow in New York City based on historical spatial and temporal data.

Key Contributions:

Data Engineering: Processed and cleaned millions of rows of NYC Taxi and Limousine Commission (TLC) data using Python (Pandas, NumPy), handling missing values, outliers, and formatting datetime features.

Feature Engineering: Extracted complex temporal features (rush hour flags, day of week) and geospatial features (pickup/drop-off coordinates, distance calculations) to improve model accuracy.

Machine Learning: Built and optimized regression models (e.g., XGBoost, Random Forest) using Scikit-Learn to predict trip durations, minimizing Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE).

Data Visualization: Created interactive geospatial visualizations using Plotly to map high-traffic density zones and bottleneck areas across Manhattan and surrounding boroughs.