A Kaggle Grandmaster cheated in $25k AI contest with hidden code

Re:A Kaggle Grandmaster cheated in $25k AI contest with hidd... by quickthrower2 on Wednesday January 22, 2020 @09:25PM (#22124268)

Wouldn’t it be smarter to use the hidden data as training data rather than hard code it in?

[ Parent | on HN ]

Re:A Kaggle Grandmaster cheated in $25k AI contest with hidd... by blurps on Wednesday January 22, 2020 @09:33PM (#22124305)

This only works when you don't win. You have to upload your code, including model training code, they won't accept a trained model binary.

[ Parent | on HN ]

Re:A Kaggle Grandmaster cheated in $25k AI contest with hidd... by MiroF on Wednesday January 22, 2020 @10:24PM (#22124544)

If they were smart, they would have used the whole set for hyperparameter tuning. That would be essentially undetectable.

[ Parent | on HN ]

Re:A Kaggle Grandmaster cheated in $25k AI contest with hidd... by jeffshek on Thursday January 23, 2020 @03:45AM (#22125932)

Not if the underlying model was bad, no tweaking of hyperparameters can change that. It's safe to say, he probably did consider this (and it probably didn't work well enough).

[ Parent | on HN ]