Back to all updates

12 months ago

Official Test Dataset Now Available

Dear Participants,

Thank you once again for being part of the AI Bounty Hackathon. We’re thrilled to see your energy and ideas already taking shape!

As shared during onboarding, the dataset provided yesterday (with Loan_Approved labels) was intended for training and model development. Today, we’re officially entering the evaluation phase of the competition.

We are now releasing the official test dataset that you will use to generate your final predictions.

Download Dataset Folder Here

We’ve attached a new file:
test.csv – unlabeled applicant records.

• Your task is to apply your trained model to this test data and predict whether each application should be Approvedor Denied.

• The Loan_Approved column has been intentionally removed to simulate real-world deployment and ensure a fair scoring process.

Please generate a file named submission.csv with:

• The ID from each row in test.csv

• Your predicted value for Loan_Approved (either Approved or Denied)

Your file should follow this format: ID, Loan_Approved

We will evaluate your submissions using a hidden answer key that contains the true outcomes for the test dataset. This ensures a fair, unbiased comparison across all submissions and discourages overfitting.

Important: Please do not use the test file for training. It is for prediction only.

Submissions should be made through Final Submission Form
OR
Devpost Project Page.

If you have any questions about the format or process, feel free to reply to this email or drop your questions in the community Slack. Join Slack Here

The HackTheFest Team