import { DataMatrix, LabelVector } from '../clustering/types';
/**
 * Formula: CH = (BSS / (k - 1)) / (WSS / (n - k))
 * where BSS = between-cluster sum of squares, WSS = within-cluster sum of squares.
 *
 * Noise (`-1`) samples are excluded before computing. When noise leaves fewer than
 * two clusters the score is a defined `0` rather than an error. Variance-based,
 * so no `metric` parameter.
 *
 * @throws Error if there are fewer than 2 clusters and no noise was present
 */
export declare function calinski_harabasz(X: DataMatrix, labels: LabelVector): number;
/** Manages tensor lifecycle manually (no tf.tidy wrapper) to bound peak memory for large datasets. */
export declare function calinski_harabasz_efficient(X: DataMatrix, labels: LabelVector): number;
//# sourceMappingURL=calinski_harabasz.d.ts.map