Summary: XGBoost and ensembles take the Kaggle cake but they’re mainly used for classification tasks. Some tools like factorization machines and vowpal wabbit make occasional appearances.
My friend, Josh Jacquet, and I competed in the DMA’s 2016 Analytics Challenge (powered by EY) and placed 4th out of the 50 entrants. Given that the majority of the other contestants were agencies vying for a little exposure, I think we did well.
Summary: Kaggle competitors spend their time exploring the data, building training set samples to build their models on representative data, explore data leaks, and use tools like Python, R, XGBoost, and Multi-Level Models.
FastML has a great breakdown of the cutting edge Artificial Intelligence algorithms. Don’t believe the hype.
Essentially that’s what it is. A company acquired by Google developed a fairly general “AI” that can play a series of Atari games. Nature published a video interview with the researchers and it’s pretty interesting. Turns out [3:40] it’s a neural network that takes input as what’s on the game screen and then looks at […]