Bias-Variance Tradeoff
Updated 2026-08-05
INTRODUCTION
English translation pending.
CORE DEFINITION
A central result in statistical learning: a model's expected error decomposes into bias, variance, and irreducible noise, and the first two cannot be minimized simultaneously. Simple models underfit with high bias, complex models overfit with high variance, and the practical goal is not to eliminate either but to find the complexity that minimizes total generalization error. The tradeoff assumes a finite sample and a model family whose capacity can be varied.
SCAFFOLDING EFFECT
Reduce cognitive load
- Complexity dial: decide whether your model or rule is too rigid or too reactive. - Error diagnosis: read training against validation error to separate underfitting from overfitting. - Stopping rule: halt added complexity at the point where validation error stops falling.
Anchor fast decisions
Bias is systematic error from a model too simple to represent the true relationship; variance is sensitivity to the particular sample it was trained on. Raising capacity lowers bias but raises variance, and lowering capacity does the reverse, so total error traces a U-shaped curve with a minimum at intermediate complexity. Past that minimum, extra fit is memorized noise that fails on new data.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- en.wikipedia.orghttps://en.wikipedia.org/wiki/Bias%E2%80%93variance_tradeoffverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS