RidgeClassifier

Algorithm overview

RidgeClassifier adapts ridge regression for classification by fitting one-vs-rest linear models and selecting the class with the highest score. It is useful for fast, interpretable classification on numeric tabular data.

JavaScript implementation

@kanaries/ml exposes Linear.RidgeClassifier with a JavaScript estimator API. It supports binary and multiclass numeric labels through one-vs-rest fitting.

Quick start example

import { Linear } from '@kanaries/ml';

const X = [[0, 0], [0, 1], [2, 2], [3, 2]];
const y = [0, 0, 1, 1];

const clf = new Linear.RidgeClassifier({ alpha: 1 });
clf.fit(X, y);
const pred = clf.predict([[1, 1], [3, 3]]);
console.log(pred);

Detailed API reference

new Linear.RidgeClassifier(props?: {
  alpha?: number;
  fitIntercept?: boolean;
})

Methods:

• fit(trainX: number[][], trainY: number[]): void
• predict(testX: number[][]): number[]

The classifier sorts numeric classes ascending and fits one RidgeRegression model per class.