Home › Glossary › Fundamentals › Empirical risk minimization (ERM)

Intermediate · Fundamentals

Empirical risk minimization (ERM)

Visual diagram · (in preparation) · Math · (in preparation) · Worked example · 3 difficulty levels.

TL;DR. A machine learning principle that aims to find a model that performs best on the observed training data.

Technical Definition

A machine learning principle that aims to find a model that performs best on the observed training data.

How it works

Empirical Risk Minimization (ERM) is a common learning strategy where the goal is to select a model that minimizes the error or loss calculated on the training dataset. This approach assumes that a model performing well on the training data will generalize well to unseen data. However, it can sometimes lead to overfitting if not balanced with regularization.

Related Concepts

  • Loss Function — A mathematical function that measures how far the model's predictions are from the actual values, guiding the learning process.
  • Overfitting — When a model learns noise and specific patterns in training data too well, causing it to perform poorly on new, unseen data.
  • Generalization — A model's ability to perform well on new, unseen data — not just its training set.

Further Reading

  • Google ML Glossary