/**
 * Matrix Utility Functions for Machine Learning
 *
 * Provides numerically stable operations for ML algorithms
 */
/**
 * Log-Sum-Exp trick for numerical stability
 *
 * Computes log(sum(exp(logValues))) in a numerically stable way
 *
 * @param logValues - Array of log values
 * @returns log(sum(exp(logValues)))
 *
 * @example
 * ```typescript
 * const result = logSumExp([-1000, -999, -1001]); // ≈ -998.59
 * ```
 */
export declare function logSumExp(logValues: number[]): number;
/**
 * Normalize rows of a matrix (each row sums to 1)
 *
 * @param matrix - Input matrix
 * @returns Matrix with normalized rows
 *
 * @example
 * ```typescript
 * const matrix = [[1, 2, 3], [4, 5, 6]];
 * const normalized = normalizeRows(matrix);
 * // [[0.167, 0.333, 0.5], [0.267, 0.333, 0.4]]
 * ```
 */
export declare function normalizeRows(matrix: number[][]): number[][];
/**
 * Normalize an array (values sum to 1)
 *
 * @param arr - Input array
 * @returns Normalized array
 *
 * @example
 * ```typescript
 * const arr = [1, 2, 3, 4];
 * const normalized = normalizeArray(arr); // [0.1, 0.2, 0.3, 0.4]
 * ```
 */
export declare function normalizeArray(arr: number[]): number[];
/**
 * Add small noise to break symmetry
 *
 * @param matrix - Input matrix
 * @param noiseLevel - Level of noise (default: 1e-4)
 * @returns Matrix with added noise
 */
export declare function addNoise(matrix: number[][], noiseLevel?: number): number[][];