HumachLearnLearn · explore · apply
Start learning
Model Building & Algorithm Learning

Algorithm Step-by-Step Mode

Slow an algorithm down and inspect the sequence of operations that is usually hidden inside a single fit or transform call.

7 · Regression8 · Clustering9 · Decision boundaries10 · Neural networks11 · Gradient descent12 · Ensembles13 · Step-by-step
Lab concept guide

What to observe while you experiment

An algorithm that looks like one call in a library is usually a sequence of state changes: choose a candidate operation, update an internal quantity, test a stopping rule, and repeat. This lab exposes that sequence so you can connect the final result to the operations that produced it.

MechanismAdvance one step at a time and record which data, parameter or model state changes at that step.
Failure modeJumping straight to the final state and assuming you understand the algorithm because the output looks plausible.
VerificationPredict the next state before clicking Step, then compare the prediction with the actual update and explain any mismatch.
Experiment deliberately
Reset, choose a tiny problem, and narrate three consecutive updates. Then change one control and predict which step in the sequence should change first.
Choose an algorithm and advance one operation at a time.

Algorithm view

What happened in this step?

State