University of Toronto Deep Learning Group won the Merck Drug Discovery Competition

A team led by George Dahl, a graduate student from University of Toronto won the Kaggle’s Merck Drug Discovery Challenge[1]. Machine Learning algorithm that they have used to won the challenge is using an ensemble of supervised learning algorithms. But the most important part of this algorithm is the deep neural network that they have used with a recent training trick called droputs [2].

[1] Kaggle’s blog (No Free Hunch: the sport of data science), Merck Competition Results - Deep NN and GPUs come out to play, 31 October 2012
http://blog.kaggle.com/2012/10/31/merck-competition-results-deep-nn-and-gpus-come-out-to-play/

[2] Kaggle’s blog (No Free Hunch: the sport of data science), Deep Learning How I Did It: Merck 1st place interview, 1 November 2012

http://blog.kaggle.com/2012/11/01/deep-learning-how-i-did-it-merck-1st-place-interview/

1 comment to University of Toronto Deep Learning Group won the Merck Drug Discovery Competition

  • I have already come.As I premosid that I would come again, I have already come. Now all the people around the world must listen to this message, I have already come