Model vs model
Contrast interpretability, complexity, data size, assumptions and evaluation metrics.
Compare algorithms → FREE • INTERACTIVE • 100% IN YOUR BROWSER
Learn Machine Learning
by seeing it work.
A visual-first machine learning tutorial that connects intuition → mathematics → algorithm → interactive experiment → real-world use. No setup, no login and no black-box demos.
15 core algorithms15 interactive labs75+ visual conceptsLocal progress
learning-loop.js
01ProblemWhat should we predict?
→
02MathWhat function fits?
↓
04VisualizeChange parameters live
←
03LearnMinimize the loss
lossconverging ✓
YOUR LEARNING STATE
Continue where you left off
GUIDED CURRICULUM
One atlas. Three ways to learn.
Start from the mathematics, start from an algorithm, or start from a real problem. Every route reconnects to the same mental map.
01
Beginner Roadmap
Follow the recommended order from data and functions to regression, classification, clustering and dimensionality reduction.
Start from zero → 02Math for ML
Connect vectors, matrices, probability, derivatives, gradients, entropy and eigenvectors directly to the models that use them.
Explore the math → 03Choose a Model
Describe the prediction problem and let the frontend decision engine narrow the model family and explain the trade-offs.
Open decision engine →MACHINE LEARNING ALGORITHMS
Explore the core models
Each tutorial includes intuition, mathematical formulation, assumptions, interactive visualization, evaluation, strengths, limitations and real-world applications.
LEARN BY DOING
The page should react when your understanding changes.
INTERACTIVE LAB
Change a parameter. Watch the mathematics respond.
Drag points in KNN, move a regression sample, change K in K-Means, grow a decision tree, or train an SVM margin. Live captions translate the visual state back into the algorithm.
Find concepts, not pages
Search terms such as gradient descent, entropy, margin, regularization or clustering.
THE BIG PICTURE
Machine learning starts with the problem type.
Supervised Learning
Learn a mapping from features to known targets.
Regression
Linear, polynomial and regularized regression.
Classification
Logistic, KNN, Naive Bayes, trees, forests and SVM.
Unsupervised Learning
K-Means, hierarchical clustering, DBSCAN and PCA.
DON'T MEMORIZE ALGORITHMS
Build a connected mental model of ML.
Start with the roadmap, experiment with every visualizer, then use the comparison and model selector to understand when each method belongs.