This project consists in making a competitive entry in the AutoML challenge. The challenge is to create a learning machine capable of learning without any human intervention (no hyper-parameter tweaking). This is not as hard as it may seem: in the third round, no entrant succeeded in making an entry that returned results on all 5 datasets, yet the simple Naive Bayes method could have won. Enter and win with simple methods!
0 Comments
|
Details
AuthorIsabelle Guyon. Chaired professor of "big data", Paris-Saclay University. President of ChaLearn.org. Archives
April 2018
Categories |