Convergence of Anisotropic Consensus-Based Optimization in Mean-Field Law

In this paper we study anisotropic consensus-based optimization (CBO), a multi-agent metaheuristic derivative-free optimization method capable of globally minimizing nonconvex and nonsmooth functions in high dimensions. CBO is based on stochastic swarm intelligence, and inspired by consensus dynamics and opinion formation. Compared to other metaheuristic algorithms like particle swarm optimization, CBO is of a simpler nature and therefore more amenable to theoretical analysis. By adapting a recently established proof technique, we show that anisotropic CBO converges globally with a dimension-independent rate for a rich class of objective functions under minimal assumptions on the initialization of the method. Moreover, the proof technique reveals that CBO performs a convexification of the optimization problem as the number of agents goes to infinity, thus providing an insight into the internal CBO mechanisms responsible for the success of the method. To motivate anisotropic CBO from a practical perspective, we further test the method on a complicated high-dimensional benchmark problem, which is well understood in the machine learning literature.

[1]  Marco Dorigo,et al.  Ant colony optimization theory: A survey , 2005, Theor. Comput. Sci..

[2]  G. Toscani,et al.  Kinetic models of opinion formation , 2006 .

[3]  Massimo Fornasier,et al.  Particle, kinetic, and hydrodynamic models of swarming , 2010 .

[4]  Jinniao Qiu,et al.  On the mean‐field limit for the consensus‐based optimization , 2021, Mathematical Methods in the Applied Sciences.

[5]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[6]  Shi Jin,et al.  A consensus-based global optimization method for high dimensional machine learning problems , 2019 .

[7]  Emile H. L. Aarts,et al.  Simulated annealing and Boltzmann machines - a stochastic approach to combinatorial optimization and neural computing , 1990, Wiley-Interscience series in discrete mathematics and optimization.

[8]  Lorenzo Pareschi,et al.  Anisotropic Diffusion in Consensus-Based Optimization on the Sphere , 2021, SIAM J. Optim..

[9]  P. Miller Applied asymptotic analysis , 2006 .

[10]  R. Pinnau,et al.  A consensus-based model for global optimization and its mean-field limit , 2016, 1604.05648.

[11]  Riccardo Poli,et al.  Particle swarm optimization , 1995, Swarm Intelligence.

[12]  Massimo Fornasier,et al.  Consensus-based optimization methods converge globally in mean-field law , 2021, ArXiv.

[13]  Ehl Emile Aarts,et al.  Simulated annealing and Boltzmann machines , 2003 .

[14]  Thomas Bäck,et al.  Evolutionary computation: Toward a new philosophy of machine intelligence , 1997, Complex..

[15]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[16]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[17]  Alex A. Freitas,et al.  Evolutionary Computation , 2002 .

[18]  Jos'e A. Carrillo,et al.  An analytical framework for a consensus-based global optimization method , 2016, 1602.00220.