TAXI_ML

Taxi_ML — A sleek Flask ML app that predicts weekly taxi riders using price, population, income, and parking data, with a modern glassmorphism UI.

Comment

Inspiration We wanted to create a tool that could help city planners, transport companies, and entrepreneurs forecast taxi demand based on real-world economic and population data — and present it in a visually appealing way that anyone could use. What it does

Taxi_ML predicts the number of weekly taxi riders given four inputs:

  • Price per week
  • Population size
  • Average monthly income
  • Average monthly parking cost

It displays the prediction instantly in a modern, glassmorphism-styled web interface with a dynamic disco light background.

How we built it

  • Dataset Preparation: Collected and formatted CSV data containing historical pricing, population, income, parking costs, and rider counts.
  • Model Training: Built and trained a regression model in Python, then exported it as model.pkl using Pickle.
  • Backend: Implemented with Flask to handle form inputs, process predictions, and render the results.
  • Frontend: HTML + CSS for a fluid UI/UX, with Google Fonts and Font Awesome icons.
  • Styling: Applied a glassmorphism theme and an animated disco gradient background.

Challenges we ran into

  • Formatting the CSV so it worked seamlessly with the ML model.
  • Designing a clean and centered output section that didn’t break the background animation.
  • Ensuring compatibility between local Python environments and deployment environments.

Accomplishments that we're proud of

  • Creating an end-to-end ML web app that is visually engaging and functional.
  • Achieving real-time predictions with minimal latency.
  • Designing a UI that is both responsive and aesthetically modern.

What we learned

  • How to integrate Flask with a trained ML model.
  • How to center and style dynamic outputs without disturbing backend logic.
  • The importance of user-friendly design in making ML tools accessible.

What's next for TAXI_ML

  • Deploying on a cloud platform (Heroku, Render, or Railway) for public access.
  • Adding data visualization charts for deeper insights.
  • Expanding dataset coverage for multiple cities and regions.
  • Enabling API access so other apps can request predictions.

Built With

Share this project:

Updates