A form of generalisation error known as Off Training Set (OTS) error was recently introduced in [Wolpert, 1996b], along with a theorem showing that small training set error does not guarantee small OTS error, unless assumptions are made about the target function. Here it is shown that the applicability of this theorem is limited to models in which the distribution generating training data has no overlap with the distribution generating test data. It is argued that such a scenario is of limited relevance to machine learning.
Cite
@article{arxiv.1912.05915,
title = {Some observations concerning Off Training Set (OTS) error},
author = {Jonathan Baxter},
journal= {arXiv preprint arXiv:1912.05915},
year = {2019}
}
Comments
Technical Report, Australian National University, August 1999