Abstract:We propose a third order dynamical system for solving a nonlinear equation in Hilbert spaces where the operator is cocoercive with respect to the solutions set. Under mild conditions on the parameters, we establish the existence and uniqueness of the generated trajectories as well as its asymptotic convergence to a solution of the equation. When the operator is strongly monotone with respect to the solutions set, we deliver an exponential convergence rate of $e^{-2t}$, which is significantly faster than the known results of second order dynamical systems. In particular, for convex optimization problems, the proposed dynamical system provides a fast convergence rate of {\bf $\mathcal{O}(\frac{1}{t^3})$} for the objective values. In addition, we discuss the applications of the proposed dynamical system to several splitting monotone inclusion problems.
What problem does this paper attempt to address?
This paper proposes a third-order dynamic system for solving nonlinear equations, where the operator is symmetric wrt the solution set. Under mild parameter conditions, the paper proves the existence and uniqueness of generating trajectories, as well as their asymptotic convergence behavior to the equation solution. When the operator is strongly monotone with respect to the solution set, the paper shows that the trajectories converge exponentially at a rate of e^(-2t). For convex optimization problems (e.g., optimization problems), the paper provides a proof of fast convergence of the objective value at a rate of O(1/t^3). In addition, the paper discusses the application of this dynamic system in several split-monotone inclusion problems.
In summary, the main contributions of this paper include:
1. Designing a third-order dynamic system to solve nonlinear equations and proving the existence and uniqueness of solutions.
2. Providing fast exponential convergence rate e^(-2t) for strongly monotone equation solutions.
3. Convergence of the objective function value for convex optimization problems at a rate of O(1/t^3).
4. Discussing the application of operators that are quasi-strong symmetric only with respect to the solution set, which is useful for dealing with split-monotone inclusion problems that are not purely residual-strong symmetric.
The structure of the paper is as follows:
1. The introduction section reviews related work and introduces the research background.
2. The preliminaries section provides definitions and lemmas, laying the foundation for subsequent analysis.
3. The existence, uniqueness, and weak convergence of the dynamic system are studied.
4. Fast exponential convergence is proved when the operator is strongly monotone with respect to the solution set.
5. Fast convergence rate O(1/t^3) of convex optimization problems is proved.
6. The application section discusses the application of the dynamic system in several split-monotone inclusion problems.
Through these results, the paper provides an effective method for handling optimization and inclusion problems with nonlinear terms, especially in guaranteeing fast convergence rates.