overfittingn.
/ˌəʊvəˈfɪtɪŋ/, U.S. /ˌoʊvərˈfɪdɪŋ/ Statistics.
The production of an analysis which corresponds too closely or exactly to a particular set of data, and may therefore fail to fit additional data or predict future observations reliably.
ΚΠ
1935 Q. Rev. Biol. 10 371/2 To us a six-variate analysis based on thirteen observations seems rather like overfitting.
1968 Philos. Trans. (Royal Soc.) A. 263 14 The longer period of analysis allows a more extended set of variables to be used without undue over-fitting, by reducing the sampling errors.
1999 Jrnl. Risk & Insurance 66 107 The danger is that models fit on so large a set of initial predictors may fail to predict insolvency through over-fitting.
This is a new entry (OED Third Edition, December 2004; most recently modified version published online March 2022).
<
n.1935