Bayes.BernoulliNB

Bernoulli Naive Bayes classifier for binary or boolean features. Continuous features can be converted to binary values using the binarize threshold. Class and feature probabilities are estimated with additive smoothing.

interface BernoulliNBProps {
    alpha?: number;
    binarize?: number | null;
    fitPrior?: boolean;
    classPrior?: number[] | null;
}
constructor(props: BernoulliNBProps = {})

Parameters

• alpha — Additive smoothing parameter applied when estimating probabilities. Defaults to 1.0.
• binarize — Threshold for binarizing input features. If null, the input is assumed to already be binary.
• fitPrior — Whether to learn class prior probabilities from the training data. When false, a uniform prior is used.
• classPrior — Optional array of prior probabilities for each class. If provided, these values override the learned priors.

Example

const clf = new BernoulliNB();
clf.fit(trainX, trainY);
const result = clf.predict(testX);

Practical guide: BernoulliNB in JavaScript and TypeScript

BernoulliNB models binary feature presence and is effective for fast text or event-flag classification baselines.

When to use BernoulliNB

• Your features represent yes/no states such as token presence, clicks, or product flags.
• You need probabilistic outputs for ranking, threshold tuning, or alert prioritization.
• You want a lightweight baseline before trying heavier linear or tree models.

Implementation workflow

1. Convert inputs to binary indicators or thresholded boolean-like features.
2. Train with fit(X, y) and inspect class probabilities with prediction outputs.
3. Tune priors and evaluate precision/recall for your target class tradeoff.

JavaScript deployment notes

• Prefer feature scaling for distance-based and gradient-based algorithms to improve stability.
• In browser apps, run heavy training in Web Workers to keep UI interactions smooth.
• Keep a simple baseline from the same module as a fallback model for comparison.

Search intents this page targets

• BernoulliNB JavaScript
• BernoulliNB TypeScript
• BernoulliNB browser machine learning
• @kanaries/ml BernoulliNB