ml

Why we included this project

For anyone doing data work in the browser, this package packs the whole mljs ecosystem behind a single script tag. Load the UMD build and you get a global ML object with the usual suspects for classical machine learning: PCA, k-means, hierarchical clustering, KNN, naive Bayes, decision trees, and random forests, plus several regression methods from simple linear to Theil-Sen. There are also small feedforward and self-organizing-map networks if you need a bit of neural flavor without a heavy framework. Since each algorithm is also published separately as an ml- prefixed npm package, Node users can install just what they need. That makes it a practical choice for dashboards, client-side data exploration, and chemometrics-style analysis where interpretable, classical algorithms beat a GPU-trained model.