Inspiration

Leveraging historical data provided by the Metropolitan Transportation Authority (MTA), our project aims to utilize existing information down to the station level, providing reasonable predictions of subway ridership across various locations in New York City. This predictive analysis may contribute to optimized resource allocation, improved service planning, and enhanced overall efficiency of the subway system. Through a data-driven approach, we seek to facilitate informed decision-making for the benefit of both commuters and the city's transportation infrastructure.

What it does

EDA and Modeling

How we built it

In EDA, we focus on the distribution map of payment methods, riders and transfers as well as the total number and change over time of different card types In Model, we use three boost models: cat boost, lgbm and Xgb and created an ensembled model of them as well as elastic net regression and random forests

Challenges we ran into

It is difficult to draw all the maps and implement and tuning the all the models.

Accomplishments that we're proud of

We succeeded in drawing the individual maps and we implemented our models

What we learned

We learned some new techniques for drawing maps and how to implement the model.s

What's next for HRT DATA ANALYSIS

  1. The accuracy of the model in the future.
  2. Adding more factor variables into our model, such as natural disasters, crime rates, major events, and economic conditions.
  3. Predicting the impact of specific events on passenger flow.

Built With

Share this project:

Updates