ALS: Augmented Lagrangian Sketching Methods for Linear Systems

TL;DR

This paper introduces two new stochastic sketching techniques, Penalty Sketching and Augmented Lagrangian Sketching, which generalize existing methods and develop a family of novel algorithms for solving linear systems with proven convergence.

Contribution

The paper develops the PS and ALS frameworks that unify and extend stochastic methods for linear systems, introducing new algorithms and convergence analysis.

Findings

Proposed methods outperform existing Sketch & Project techniques in experiments.

Established global convergence rates for the new stochastic algorithms.

Reformulated linear systems as stochastic optimization problems for enhanced analysis.

Abstract

We develop two fundamental stochastic sketching techniques; Penalty Sketching (PS) and Augmented Lagrangian Sketching (ALS) for solving consistent linear systems. The proposed PS and ALS techniques extend and generalize the scope of Sketch & Project (SP) method by introducing Lagrangian penalty sketches. In doing so, we recover SP methods as special cases and furthermore develop a family of new stochastic iterative methods. By varying sketch parameters in the proposed PS method, we recover novel stochastic methods such as Penalty Newton Descent, Penalty Kaczmarz, Penalty Stochastic Descent, Penalty Coordinate Descent, Penalty Gaussian Pursuit, and Penalty Block Kaczmarz. Furthermore, the proposed ALS method synthesizes a wide variety of new stochastic methods such as Augmented Newton Descent, Augmented Kaczmarz, Augmented Stochastic Descent, Augmented Coordinate Descent, Augmented Gaussian Pursuit, and Augmented Block Kaczmarz into one framework. Moreover, we show that the developed PS and ALS frameworks can be used to reformulate the original linear system into equivalent stochastic optimization problems namely the Penalty Stochastic Reformulation and Augmented Stochastic Reformulation. We prove global convergence rates for the PS and ALS methods as well as sub-linear $\mathcal{O}(\frac{1}{k})$ rates for the Cesaro average of iterates. The proposed convergence results hold for a wide family of distributions of random matrices, which provides the opportunity of fine-tuning the randomness of the method suitable for specific applications. Finally, we perform computational experiments that demonstrate the efficiency of our methods compared to the existing SP methods.