← Apps ← My Digital Space

Resampling Lab

Visualize how ML models are evaluated

Machine Learning Fundamentals

Why do we resample data?

A model that memorizes its training data isn't really learning. Resampling techniques divide your dataset into training and evaluation subsets so you can estimate how well a model will generalize to unseen data. Pick a method below and watch it work, step by step.

Choose a method

Dataset Overview

Click "Next Step" to start resampling

Train Test OOB
Class A Class B Class C = class centroid

Parameters

Train ratio 0.80
48 train 12 test

Actions

Progress 0 / 0

Latest Result

Train size
Test size
This step
Running mean

Accuracy Across Iterations

Each bar shows the model's accuracy on that iteration's test set · dashed line is the running mean

Final estimate
Run a resampling method to see accuracy per iteration
1.0 0.5 0.0

Hold-Out Method

Trade-offs