Abstract:Instead of minimizing the sum of all $n$ squared residuals as the classical least squares (LS) does, Rousseeuw (1984) proposed to minimize the sum of $h$ ($n/2 \leq h < n$) smallest squared residuals, the resulting estimator is called least trimmed squares (LTS). The idea of the LTS is simple but its computation is challenging since no LS-type analytical computation formula exists anymore. Attempts had been made since its presence, the feasible solution algorithm (Hawkins (1994)), fastlts.f (Rousseeuw and Van Driessen (1999)), and FAST-LTS (Rousseeuw and Van Driessen (2006)), among others, are promising approximate algorithms. The latter two have been incorporated into R function ltsReg by Valentin Todorov. These algorithms utilize combinatorial- or subsampling- approaches. With the great software accessibility and fast speed, the LTS, enjoying many desired properties, has become one of the most popular robust regression estimators across multiple disciplines. This article proposes analytic approaches -- employing first-order derivative (gradient) and second-order derivative (Hessian matrix) of the objective function. Our approximate algorithms for the LTS are vetted in synthetic and real data examples. Compared with ltsReg -- the benchmark in robust regression and well-known for its speed, our algorithms are comparable (and sometimes even favorable) with respect to both speed and accuracy criteria. Other major contributions include (i) originating the uniqueness and the strong and Fisher consistency at empirical and population settings respectively; (ii) deriving the influence function in a general setting; (iii) re-establishing the asymptotic normality (consequently root-n consistency) of the estimator with a neat and general approach.

Square-root RTS smoothing algorithms

New square-root smoothing algorithms

Kalman smoothing and block tridiagonal systems: new connections and numerical stability results

Minimum Error Entropy Rauch-Tung-Striebel Smoother

Numerically robust square root implementations of statistical linear regression filters and smoothers

Windowing and random weighting based cubature RTS smoothing for target tracking

The Invariant Rauch-Tung-Striebel Smoother

Numerically Robust Fixed-Point Smoothing Without State Augmentation

Strong stability preserving two-step Runge-Kutta methods

Fast and accurate approximation algorithms for computing floating point square root

MATLAB-based general approach for square-root extended-unscented and fifth-degree cubature Kalman filtering methods

New algorithms for computing the least trimmed squares estimator

Algorithms for Square Root of Semi-Infinite Quasi-Toeplitz M-Matrices

Inexactness and Correction of Floating-Point Reciprocal, Division and Square Root

Advanced-Step Real-time Iterations with Four Levels -- New Error Bounds and Fast Implementation in acados

Strong Tracking Adaptive Square-Root Cubatu re Kal man Filter Algorith m

Faster Quantum Algorithms with "Fractional''-Truncated Series

A Refined Harmonic Rayleigh-Ritz Procedure and an Explicitly Restarted Refined Harmonic Arnoldi Algorithm

Fast Compensated Algorithms for the Reciprocal Square Root, the Reciprocal Hypotenuse, and Givens Rotations

Design of High-Throughput Fixed-Point Complex Reciprocal/Square-Root Unit