NYC-Taxi-Demand-Prediction

• Built a Regression Tree based taxi demand prediction model to predict the number of taxis at a given space and time in NYC with 90% Rsquare performance
• Enhanced model's performance by adding novel taxi-demand relevant features from multiple open source datasets. Identified demand drivers using feature importance maps

Code can not be uploaded yet. But, some visualizations from the project could be seen below

Number of Pickup heatmap

Correlation between variables

Feature importance