How does increasing model complexity affect bias and variance?
Updated May 15, 2026
Short answer
Increasing model complexity reduces bias but increases variance.
Deep explanation
As model complexity increases, the model gains flexibility to capture complex patterns, reducing bias. However, it also becomes more sensitive to fluctuations in training data, increasing variance. This tradeoff is central to model selection and is often visualized as a U-shaped error curve.
Real-world example
Increasing depth of a decision tree improves training accuracy but may reduce test performance.
Common mistakes
- Assuming more complexity always improves performance.
Follow-up questions
- What is the optimal model complexity?
- How do you measure variance practically?