NeuroNomixer logoNeuroNomixer
  • Home
  • Blog
  • Visual Guides
  • Authors
  • Contact
Sign InSign Up
NeuroNomixer logo
HomeBlogAuthorsContactPrivacy Policy

© 2026 NeuroNomixer — Built with Next.js & Tailwind CSS

Visual Guides/Overfitting vs Underfitting Playground
Machine Learning

Overfitting vs Underfitting Playground

Fit polynomials of increasing degree to data and watch the model go from too simple to memorizing noise. Observe how training error and test error diverge when complexity grows unchecked.

Good Fit

Balanced bias-variance, generalizes well

Regimes ObservedUnderfittingGood FitOverfitting

Sign in to save your progress.

Train
Test (hidden from model)
degree 2

Error Comparison

Train MSE0.0625
Test MSE0.0424
Test/Train ratio: 0.68×

Dataset

Polynomial Degree

2
1 (linear)12 (overfit)

Training Points

15

Noise Level

0.20

Key Insight

When test error rises while train error falls, the model is memorizing noise instead of learning the pattern. The gap between the two errors measures generalization.

← Previous GuideNext Guide →