project-3 runs as a Kaggle competition
Kaggle is popular platform for data science competitions. Project 3 was designed to ensure that all students understand trade-offs involved in creating classification models, and complexities surrounding working with imbalanced data.
You need to:
- create a new Kaggle account.
- register for a competition using the following link: https://www.kaggle.com/t/fc5c11853b2d49df9b65bc6837965c05
- send me your id, to ensure that every participant has only one account.
This competition will end on the 27st of January, and we will be covering a lot of material required to succeed in this competition in the coming weeks. We will also overview how you can submit your predictions to Kaggle. You will be required to send me a link to your private GitHub Enterprise repository containing the code that you used to produce your predictions by the 31st of January.
Your overall performance for this project would be determined by the combination of your best private leaderboard score and amount of effort you invested in producing it.
There are no teams in this competition, and everyone needs to submit their own work. However, I encourage everyone to discuss interesting approaches for solving this problem, as long as you aren’t sharing your solution code.