Back to results for “Machine Learning Fundamentals”
Machine Learning Fundamentals › Bias-Variance Tradeoff

How does Bias-Variance Tradeoff work

Beginner

Answer

A model with high bias makes strong assumptions and underfits the data, while a model with high variance follows the training data very closely and overfits. The goal is to find a balance where both errors are reasonably low.
Real-world example A simple straight line model predicting house prices may miss important patterns, high bias, while a very complex model may fit every small detail of the training houses and fail on new houses, high variance.

Common follow-ups: What is overfitting, What is regularization, What is model complexity