About
This site is intended to serve as a common resource for Machine Learning algorithms implemented in JavaScript and run in a Web browser.
Data Mining: Practical Machine Learning Tools and Techniques
Witten, Frank, Hall, Pal & Foulds · 5th Edition · University of Waikato
Most algorithms implemented here follow the descriptions in this book. As a result, data analysis tools are available which generate useful statistics about datasets loaded into the browser — including attribute distributions, missing value counts, and evaluation metrics.
What it does
Load any ARFF or CSV dataset and the Explorer gives you four tabs:
Inspect attribute types, statistics, class distribution, and missing value counts.
Train k-NN or Naïve Bayes. Evaluate using cross-validation or a percentage hold-out split. See accuracy, confusion matrix, and per-class precision/recall.
Run k-Means with configurable k. Visualize cluster centroids and within-cluster sum of squares.
Plot any two numeric attributes against each other. Points are coloured by class label.
Tech stack
Open source
MIT licensed. Contributions welcome.
View on GitHubmachinelearning.js.org · open source · MIT · Marin's Web Site