The Analysis of Optimization Algorithms, A Dissipativity Approach

Abstract

Optimization problems in engineering and applied mathematics are typically solved in an iterative fashion, by systematically adjusting the variables of interest until an adequate solution is found. The iterative algorithms that govern these systematic adjustments can be viewed as a control system. In control systems, the output in measured and the input is adjusted using feedback to drive the error to zero. Similarly, in iterative algorithms, the optimization objective is evaluated and the candidate solution is adjusted to drive it toward the optimal point. Choosing an algorithm that works well for a variety of optimization problems is akin to robust controller design. Just as dissipativity theory can be used to analyze the stability properties of control systems, it can also be used to analyze the convergence properties of iterative algorithms. By defining an appropriate notion of "energy" that dissipates with every iteration of the algorithm, the convergence properties of the algorithm can be characterized. This article formalizes the connection between iterative algorithms and control systems and shows through examples how dissipativity theory can be used to analyze the performance of many classes of optimization algorithms. This control-theoretic viewpoint enables the selection and tuning of optimization algorithms to be performed in an automated and systematic way.

References in corpus (5)

• Optimal parameter selection for the alternating direction method of multipliers (ADMM): quadratic problems
• Analysis and Design of Optimization Algorithms via Integral Quadratic Constraints
• A General Analysis of the Convergence of ADMM
• A Unified Analysis of First-Order Methods for Smooth Games via Integral Quadratic Constraints
• A Unified Analysis of Stochastic Optimization Methods Using Jump System Theory and Quadratic Constraints

Cited by in corpus (7)

• Towards a Systems Theory of Algorithms
• Quantum computing through the lens of control: A tutorial introduction
• Time-Varying Convex Optimization: A Contraction and Equilibrium Tracking Approach
• Automated tight Lyapunov analysis for first-order methods
• Analysis of Gradient Descent with Varying Step Sizes using Integral Quadratic Constraints
• Bringing Quantum Systems under Control: A Tutorial Invitation to Quantum Computing and Its Relation to Bilinear Control Systems
• Accelerated Gradient Methods for Nonconvex Optimization: Escape Trajectories From Strict Saddle Points and Convergence to Local Minima