Inteview: Discover the Methodology and Mindset of a Kaggle Master
Last Updated on July 5, 2019
What does it take to do well in competitive machine learning? To really dig into this question, you need to dig into the people that do well.
In 2010 I participated in a Kaggle competition to predict the outcome of chess games in the future. It was a fascinating problem because it required you to model the rating of the players from historical games and propagating those ratings into the future to make predictions.
I did so-so on the competition (17th or 10%), but Diogo Ferreira took 4th place. Diogo is a fascinating guy and an excellent communicator, and in 2011 I had the opportunity to interview him about his participation in the competition. This post is an edited version of that interview.
The interview is divided into four parts:
- Part 1 explores Diogo’s background and methodology for problem solving
- Part 2 on the Kaggle chess ratings competition
- Part 3 dives into Diogo’s solution to the chess ratings competition
- Part 4 leaves us with some final thoughts for being competitive machine learning practitioners (best part!)
Grab a drink, sit back and Enjoy!
To finish reading, please visit source site