Skip to content

johri-lab/MTA-Commute-Pal

 
 

Repository files navigation

MTA Commute Pal - Columbia Data Science Society Datathon Fall 2023

Visualization Links

Subway Commute made Better.


A Preview of What Commute Pal Is:

Landing page:

Home Screenshot

Heatmap of 24 hours of live Rider Count

Rider Count HeatMap

Marker Map of Most Popular Payment Methods at a Station

Popular Payment Marker Map

Heat Map of Stations using a particular Payment method

Metro Card Usage Heatmap

The problem MTA Commute Pal solves

The New York City subway system is one of the busiest in the world, with over 5 million riders per day. However, it is also one of the most congested, with delays and overcrowding a common occurrence. This is due in part to the difficulty of predicting periods of increased ridership, which can lead to overcrowding and delays.

Solution

We propose to develop a system that combines the MTA dataset, geographical patterns of temperature, different card types stacking at each station, and traffic footfall to predict periods of increased ridership. This system will use machine learning to identify patterns in the data that can be used to forecast future ridership levels.

Features offered by MTA Commute Pal

  • Reduced congestion

    • By predicting periods of increased ridership, the MTA can run more trains at those times to help prevent overcrowding.
  • Improved service to handle footfall prediction

    • By understanding how ridership patterns vary over time and space, the MTA can improve its service planning and delivery (effective labor resources)
  • Increased customer satisfaction

    • By reducing congestion and improving service, our system will make the subway a more pleasant and reliable experience for riders. One trip at a time.

Technology Stack and Dependencies

  • Time Series Prediction
  • Pandas
  • Seaborn
  • Numpy
  • Python
  • HTML
  • CSS
  • JavaScript

Thank You!

Contributors

Shikhar Johri

Shikhar Johri

Shivam Shekhar

Shivam Shekhar

Mohsin Chougale


Mohsin Chougale

About

Columbia University Data Science Hackathon Project 2023

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 79.0%
  • HTML 12.7%
  • Python 4.1%
  • JavaScript 3.8%
  • CSS 0.4%