The bias–variance tradeoff describes the relationship between errors caused by overly simple assumptions and errors caused by sensitivity to training data. Models with high bias may underfit by missing important patterns, while models with high variance may overfit by capturing noise. Effective model design seeks a balance that generalizes well to unseen data.
In statistics and machine learning, the bias–variance tradeoff is the property of a set of predictive models whereby models with a lower bias in parameter estimation have a higher variance of the parameter estimates across samples, and vice versa.