Label Propagation in JavaScript with @kanaries/ml

Learn what Label Propagation does, when to use it, and how to run LabelPropagation in JavaScript or TypeScript with @kanaries/ml for browser and Node.js applications.

Algorithm overview

LabelPropagation spreads labels through similarity graphs to leverage unlabeled data in transductive settings.

This algorithm is especially useful when:

Only a small subset of samples is labeled.
Similarity graph structure reflects class continuity.
You need to bootstrap labels before training a final supervised model.

@kanaries/ml exposes Label Propagation in JavaScript for semi-supervised workflows where only a small portion of the dataset is labeled. This is useful for browser-based data tools, internal review apps, or Node.js services that want to spread labels through a similarity graph before downstream supervised training.

Keeping the algorithm in JS makes it easier to combine labeling interfaces, graph construction, and model-assisted annotation inside one product experience.

Quick start

LabelPropagation in Python vs JavaScript / TypeScript

If you searched for "LabelPropagation in JavaScript" or "LabelPropagation in TypeScript", this section maps the familiar scikit-learn call to the equivalent @kanaries/ml usage for browser and Node.js runtimes.

Python

scikit-learn

from sklearn.semi_supervised import LabelPropagation

X = [[0, 0], [0.1, 0.2], [1, 1], [1.1, 0.9]]
y = [0, -1, 1, -1]

model = LabelPropagation()
model.fit(X, y)
pred = model.predict(X)

JavaScript / TypeScript

@kanaries/ml

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

const X = [[0, 0], [0.1, 0.2], [1, 1], [1.1, 0.9]];
const y = [0, -1, 1, -1];

const model = new SemiSupervised.LabelPropagation();
model.fit(X, y);
const pred = model.predict(X);

Quick JavaScript example

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

const X = [[0, 0], [0.1, 0.2], [1, 1], [1.1, 0.9]];
const y = [0, -1, 1, -1];

const model = new SemiSupervised.LabelPropagation();
model.fit(X, y);
const pred = model.predict(X);

Detailed API reference

interface LabelPropagationOptions {
    kernel?: 'rbf' | 'knn' | ((X: number[][], Y: number[][]) => number[][]);
    gamma?: number;
    nNeighbors?: number;
    maxIter?: number;
    tol?: number;
}
constructor(options: LabelPropagationOptions = {})

Label propagation assigns labels to unlabeled data by propagating information from labeled points across a graph defined by a kernel.

Methods

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

Implementation workflow

Build feature representations and seed reliable initial labels.
Fit LabelPropagation and inspect propagated label confidence.
Validate on known labels and tune graph-related hyperparameters.

JavaScript deployment notes

Use Label Propagation when the similarity graph is meaningful and labels can reasonably spread through neighborhoods.
Validate propagated labels on a trusted subset before using them downstream.
It works best as a label-bootstrapping step, not as a replacement for careful annotation strategy.

Label Propagation