Flight Rating RAMP Challenge

This repository contains the Flight Rating RAMP challenge, designed as a structured competition to evaluate predictive modeling solutions for a specific dataset. Created for use with RAMP (Rapid Analytics and Model Prototyping), this challenge framework is ready to be deployed on RAMP Studio, where participants can collaborate on predictive modeling tasks and benchmark their solutions.

Project Structure

The project is organized across two main directories:

• ramp-data/flight-rating: Contains data preparation scripts and dependencies to set up and clean the dataset.

  • data/: Directory to store train/test data files.
  • prepare_data.py: Script for data preparation.
  • requirements.txt: Lists all necessary dependencies.

• ramp-kits/flight-rating: The primary kit directory, containing files needed to configure and launch the challenge.

  • data/: Placeholder for challenge data.
  • submissions/starting_kit/: Includes a baseline submission to help participants get started.
  • flight-rating_starting_kit.ipynb: Jupyter notebook with an overview of the problem, exploratory analysis, and example solution.
  • problem.py: Defines the predictive task using RAMP workflow specifications.
  • environment.yml and requirements.txt: Dependency files for conda and pip setup.

This structure is intended to provide a clear, organized approach for participants, ensuring they can easily prepare, understand, and participate in the challenge on RAMP Studio.