What is Overfitting in FRM Part 1?

Fitting noise or sample-specific detail so closely that performance deteriorates on new data. A large training-versus-test performance gap is a classic warning; regularization and validation can help.

Overfitting

FRM Part 1 glossary · Quantitative Analysis

Definition

Fitting noise or sample-specific detail so closely that performance deteriorates on new data.

Overfitting in plain English

A large training-versus-test performance gap is a classic warning; regularization and validation can help.

Practice Overfitting

Within Quant — Machine Learning Methods, which term matches this definition: “learning sample noise so closely that…

Related FRM Part 1 terms

Bayes' theorem
Bootstrapping (term structure)
Central limit theorem
Copula
Exponentially weighted moving average (EWMA)
Generalized Autoregressive Conditional Heteroskedasticity (GARCH)
Heteroskedasticity
Monte Carlo simulation
Ordinary least squares (OLS)
p-value

FRM Part 1 practice questions · FRM Part 1 flashcards · FRM Part 1 study guide