ElasticNet

Algorithm overview

ElasticNet combines L1 and L2 penalties. It is useful when you want sparse-ish linear models but pure Lasso is too unstable, especially with correlated features.

JavaScript implementation

@kanaries/ml provides Linear.ElasticNet for browser and Node.js regression workflows. l1Ratio controls the balance between Ridge-like and Lasso-like behavior.

Quick start example

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

const X = [[0, 1], [1, 1], [2, 0], [3, 0]];
const y = [1, 2, 3, 4];

const model = new Linear.ElasticNet({ alpha: 0.1, l1Ratio: 0.5 });
model.fit(X, y);
const pred = model.predict([[4, 0]]);

Detailed API reference

new Linear.ElasticNet(props?: {
  alpha?: number;
  l1Ratio?: number;
  fitIntercept?: boolean;
  maxIter?: number;
  tol?: number;
})

Options:

• alpha?: number, default 1.
• l1Ratio?: number, default 0.5. Must be between 0 and 1.
• fitIntercept?: boolean, default true.
• maxIter?: number, default 1000.
• tol?: number, default 1e-6.

Methods:

• fit(X: number[][], Y: number[]): void
• predict(X: number[][]): number[]

When l1Ratio is 0, the implementation delegates to Ridge regression. When l1Ratio is 1, it delegates to Lasso regression.