MultinomialNB

Algorithm overview

MultinomialNB is a naive Bayes classifier for non-negative feature counts, such as token counts or frequency-style encoded data.

JavaScript implementation

@kanaries/ml provides Bayes.MultinomialNB for JavaScript and TypeScript applications that need fast probabilistic classification without leaving the browser or Node.js runtime.

Quick start example

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

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

const clf = new Bayes.MultinomialNB({ alpha: 1.0 });
clf.fit(X, y);
const pred = clf.predict([[1, 0, 1], [0, 2, 2]]);

Detailed API reference

new Bayes.MultinomialNB(props?: {
  alpha?: number;
  forceAlpha?: boolean;
  fitPrior?: boolean;
  classPrior?: number[] | null;
})

Methods:

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

Feature values must be non-negative.