10 min listen
Thresholdout: Down with Overfitting
ratings:
Length:
16 minutes
Released:
Nov 27, 2015
Format:
Podcast episode
Description
Overfitting to your training data can be avoided by evaluating your machine learning algorithm on a holdout test dataset, but what about overfitting to the test data? Turns out it can be done, easily, and you have to be very careful to avoid it. But an algorithm from the field of privacy research shows promise for keeping your test data safe from accidental overfitting
Released:
Nov 27, 2015
Format:
Podcast episode
Titles in the series (100)
Hunting for the Higgs: Machine learning and particle physics go together… by Linear Digressions