Hacker News new | ask | show | jobs
by hntrader 1902 days ago
Delusional - absolutely. Such a large proportion fall into the trap of iteratively overfitting to the test set without much of a care for the basics such as data quality, what question to ask, etc. It's a problem.