Concept

Avoiding Test Set Decisions During Regular Progress Tracking

A team may evaluate the system on the test set regularly to track progress, such as weekly or monthly. The test set should not be used to make algorithm decisions, including whether to roll back to a previous system. Using the test set for such decisions can cause overfitting to the test set, so the test set can no longer be counted on to give a completely unbiased estimate of system performance.

0

1

Updated 2026-05-26

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Machine Learning Strategy

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Learn After