This is the code for a group project I participated in which was all about classification. For our Dataset we chose to work with the OkCupid dataset used in the paper by Emil Kirkegaard. It's a very interesting dataset to work with, and certainly comes with some associated challanges. I did a write up on the project on my blog if reading more is your thing.
If you are interested in reproducing the results or using the code as a jumping off point for your own exploration, you'll need to run two of the notebooks, BuildDatabase.ipynb and DataMunging.ipynb in that order. That will provide you with the data conveniently stored in a sqLite database which makes the exploration a little easier.
Additionally there is a flask app for playing around with the results.
Dating site data is super interesting.
Most awesome contributors include:
Bob Tian
Daniel Vigel
Ron Olshansky-Lucero