MachineLearning.js

Explorer Docs About GitHub

Machine Learning in JavaScript

A JavaScript reimagining of Weka — load a dataset, train a model, and read the results. Nothing to install. Everything runs locally.

Open Explorer Witten & Frank Book

k-NN Naïve Bayes k-Means K2 BN Builder

Zero Install

Preprocess

Upload CSV or ARFF datasets. Inspect attribute statistics, distributions, and missing values instantly.

Classify

Run k-NN and Naïve Bayes classifiers. Evaluate with cross-validation or a percentage hold-out split.

Cluster

Discover structure with k-Means. Visualize cluster assignments and within-cluster variance.

Visualize

Explore your data with interactive 2-D scatter plots — select any two attributes and colour by class.

BN Builder

Build and query Bayesian Networks interactively. Set CPTs, observe evidence, run exact inference.

Inspired by Weka 3 from the University of Waikato and the book Data Mining: Practical Machine Learning Tools and Techniques.

Inspired by Weka from the University of Waikato · Marin's Web Site

For educational and research use only. Results are provided “as is” without warranty of any kind and should not be relied upon for clinical, financial, safety-critical, or other consequential decisions.