Out-of-bag evaluation Machine Learning
noun phrase
Definition: An internal evaluation method for bootstrap-based ensemble models (such as random forests) in which each training Example in context: is evaluated using only the trees that were not trained on that Example in context:, yielding an out-of-bag estimate of prediction error or score [scikit-learn Documentation].
Example in context: “As the model is based on improving the performances of week learners and making a final decision using averaging or majority voting, the RFR use the «out-of-bag: OOB» for quantifying the calculted error and for ranking the variables in terms of importance using the permutation strategy.” [Heddam et al. 2024]
Synonym: OOB evaluation; out-of-bag estimation; OOB estimate
Related: bootstrap aggregation (bagging); random forest; OOB score; OOB error; bootstrap sample; permutation importance