clustering-tfjs/dist/validation/silhouette.d.ts

High-performance TypeScript clustering algorithms (K-Means, Spectral, Agglomerative) with TensorFlow.js acceleration and scikit-learn compatibility

import { DataMatrix, LabelVector } from '../clustering/types';
/**
 * Computes the Silhouette score.
 *
 * The silhouette coefficient for a sample is (b - a) / max(a, b) where:
 * - a is the mean distance between a sample and all other points in the same cluster
 * - b is the mean distance between a sample and all points in the nearest cluster
 *
 * The score ranges from -1 to +1:
 * - +1: Sample is far from neighboring clusters (well clustered)
 * - 0: Sample is on or very close to the decision boundary
 * - -1: Sample might have been assigned to the wrong cluster
 *
 * @param X - Data matrix of shape [n_samples, n_features]
 * @param labels - Cluster labels for each sample
 * @returns The mean silhouette score across all samples
 * @throws Error if k <= 1
 */
export declare function silhouetteScore(X: DataMatrix, labels: LabelVector): number;
/**
 * Computes the Silhouette score for specific samples (subset).
 * Useful for large datasets where computing all pairwise distances is prohibitive.
 *
 * @param X - Data matrix of shape [n_samples, n_features]
 * @param labels - Cluster labels for each sample
 * @param sampleIndices - Indices of samples to compute silhouette for
 * @returns The mean silhouette score for the specified samples
 */
export declare function silhouetteScoreSubset(X: DataMatrix, labels: LabelVector, sampleIndices: number[]): number;
//# sourceMappingURL=silhouette.d.ts.map