StoMADS: Stochastic blackbox optimization using probabilistic estimates

This work introduces StoMADS, a stochastic variant of the mesh adaptive direct-search (MADS) algorithm originally developed for deterministic blackbox optimization. StoMADS considers the unconstrained optimization of an objective function f whose values can be computed only through a blackbox corrupted by some random noise following an unknown distribution. The proposed method is based on an algorithmic framework similar to that of MADS and uses random estimates of function values obtained from stochastic observations since the exact deterministic computable version of f is not available. Such estimates are required to be accurate with a sufficiently large but fixed probability and satisfy a variance condition. The ability of the proposed algorithm to generate an asymptotically dense set of search directions is then exploited to show convergence to a Clarke stationary point of f with probability one, using martingale theory.

[1]  R. Durrett.  Probability: Theory and Examples , 1993 .

[2]  Raghu Pasupathy,et al.  ASTRO-DF: A Class of Adaptive Sampling Trust-Region Algorithms for Derivative-Free Stochastic Optimization , 2016, SIAM J. Optim..

[3]  F. Clarke.  Optimization And Nonsmooth Analysis , 1983 .

[4]  Michael C. Ferris,et al.  A Direct Search Algorithm for Optimization with Noisy Function Evaluations , 2000, SIAM J. Optim..

[5]  Jorge J. Moré,et al.  Digital Object Identifier (DOI) 10.1007/s101070100263 , 2001 .

[6]  Charles Audet,et al.  Mesh Adaptive Direct Search Algorithms for Constrained Optimization , 2006, SIAM J. Optim..

[7]  A Stochastic Line Search Method with Convergence Rate Analysis , 2018, 1807.07994.

[8]  J. Kiefer,et al.  Stochastic Estimation of the Maximum of a Regression Function , 1952 .

[9]  Jeffrey Larson,et al.  Stochastic derivative-free optimization using a trust region framework , 2016, Comput. Optim. Appl..

[10]  Stefan M. Wild,et al.  Benchmarking Derivative-Free Optimization Algorithms , 2009, SIAM J. Optim..

[11]  John A. Nelder,et al.  A Simplex Method for Function Minimization , 1965, Comput. J..

[12]  Katya Scheinberg,et al.  Global convergence rate analysis of unconstrained optimization methods based on probabilistic models , 2015, Mathematical Programming.

[13]  Katya Scheinberg,et al.  Convergence Rate Analysis of a Stochastic Trust-Region Method via Supermartingales , 2016, INFORMS Journal on Optimization.

[14]  Charles Audet,et al.  Reducing the Number of Function Evaluations in Mesh Adaptive Direct Search Algorithms , 2012, SIAM J. Optim..

[15]  Charles Audet,et al.  A Survey on Direct Search Methods for Blackbox Optimization and their Applications , 2014 .

[16]  Katya Scheinberg,et al.  Stochastic optimization using a trust-region method and random models , 2015, Mathematical Programming.

[17]  Charles Audet,et al.  The Mesh Adaptive Direct Search Algorithm for Granular and Discrete Variables , 2018, SIAM J. Optim..

[18]  Rui Shi,et al.  A Stochastic Trust Region Algorithm Based on Careful Step Normalization , 2017, INFORMS J. Optim..

[19]  J. S. Ivey,et al.  Nelder-Mead simplex modifications for simulation optimization , 1996 .

[20]  Jack P. C. Kleijnen,et al.  An Asymptotic Test of Optimality Conditions in Multiresponse Simulation Optimization , 2012, INFORMS J. Comput..

[21]  Krishnakumar Balasubramanian,et al.  Zeroth-Order Nonconvex Stochastic Optimization: Handling Constraints, High Dimensionality, and Saddle Points , 2018, Foundations of Computational Mathematics.

[22]  Charles Audet,et al.  Dynamic scaling in the mesh adaptive direct search algorithm for blackbox optimization , 2014, Optimization and Engineering.

[23]  Y. Marzouk,et al.  A trust-region method for derivative-free nonlinear constrained stochastic optimization , 2017, 1703.04156.

[24]  Julien Mairal,et al.  Estimate Sequences for Stochastic Composite Optimization: Variance Reduction, Acceleration, and Robustness to Noise , 2019, J. Mach. Learn. Res..

[25]  Katya Scheinberg,et al.  Convergence of Trust-Region Methods Based on Probabilistic Models , 2013, SIAM J. Optim..

[26]  Nikolaos V. Sahinidis,et al.  Simulation optimization: a review of algorithms and applications , 2014, 4OR.

[27]  Rui Shi,et al.  A Stochastic Trust Region Algorithm , 2017 .

[28]  Sébastien Le Digabel,et al.  Algorithm xxx : NOMAD : Nonlinear Optimization with the MADS algorithm , 2010 .

[29]  Xiaoyu Wang,et al.  Stochastic Trust-Region Methods with Trust-Region Radius Depending on Probabilistic Models , 2019, Journal of Computational Mathematics.

[30]  Charles Audet,et al.  Derivative-Free and Blackbox Optimization , 2017 .

[31]  Rabi Bhattacharya,et al.  A basic course in probability theory , 2007 .

[32]  Katya Scheinberg,et al.  Introduction to derivative-free optimization , 2010, Math. Comput..

[33]  Nicholas I. M. Gould,et al.  CUTEst: a Constrained and Unconstrained Testing Environment with safe threads for mathematical optimization , 2013, Computational Optimization and Applications.

[34]  Sébastien Le Digabel,et al.  Use of quadratic models with mesh-adaptive direct search for constrained black box optimization , 2011, Optim. Methods Softw..

[35]  Charles Audet,et al.  Parallel Space Decomposition of the Mesh Adaptive Direct Search Algorithm , 2007, SIAM J. Optim..

[36]  Pierre-Yves Bouchet.  Optimisation de boîtes noires à précision variable , 2019 .

[37]  Charles Audet,et al.  Robust optimization of noisy blackbox problems using the Mesh Adaptive Direct Search algorithm , 2016, Optim. Lett..

[38]  Charles Audet,et al.  Analysis of Generalized Pattern Searches , 2000, SIAM J. Optim..

[39]  Kuo-Hao Chang,et al.  Stochastic Nelder-Mead simplex method - A new globally convergent direct search method for simulation optimization , 2012, Eur. J. Oper. Res..

[40]  Charles Audet,et al.  OrthoMADS: A Deterministic MADS Instance with Orthogonal Directions , 2008, SIAM J. Optim..