This paper presents a stochastic modification of a limited memory BFGS method to solve bound-constrained global minimization problems with a differentiable cost function with no further smoothness. The approach is a s...This paper presents a stochastic modification of a limited memory BFGS method to solve bound-constrained global minimization problems with a differentiable cost function with no further smoothness. The approach is a stochastic descent method where the deterministic sequence, generated by a limited memory BFGS method, is replaced by a sequence of random variables. To enhance the performance of the proposed algorithm and make sure the perturbations lie within the feasible domain, we have developed a novel perturbation technique based on truncating a multivariate double exponential distribution to deal with bound-constrained problems;the theoretical study and the simulation of the developed truncated distribution are also presented. Theoretical results ensure that the proposed method converges almost surely to the global minimum. The performance of the algorithm is demonstrated through numerical experiments on some typical test functions as well as on some further engineering problems. The numerical comparisons with stochastic and meta-heuristic methods indicate that the suggested algorithm is promising.展开更多
Interior-point methods(IPMs) for linear programming(LP) are generally based on the logarithmic barrier function. Peng et al.(J. Comput. Technol. 6: 61–80, 2001) were the first to propose non-logarithmic kernel functi...Interior-point methods(IPMs) for linear programming(LP) are generally based on the logarithmic barrier function. Peng et al.(J. Comput. Technol. 6: 61–80, 2001) were the first to propose non-logarithmic kernel functions(KFs) for solving IPMs. These KFs are strongly convex and smoothly coercive on their domains.Later, Bai et al.(SIAM J. Optim. 15(1): 101–128, 2004) introduced the first KF with a trigonometric barrier term. Since then, no new type of KFs were proposed until 2020, when Touil and Chikouche(Filomat. 34(12):3957–3969, 2020;Acta Math. Sin.(Engl. Ser.), 38(1): 44–67, 2022) introduced the first hyperbolic KFs for semidefinite program(ming(SD)P). They( establishe)d that the iteration complexities of algorithms based on their proposed KFs are O(n2/3log(n/ε) and O(n3/4log(n/ε)) for large-update methods, respectively. The aim of this work is to improve the complexity result for large-update method. In fact, we present a new parametric KF with a hyperbolic barrier term. By simple tools, we show that the worst-case iteration complexity of our algorithm for the large-update method is O(√n log n log(n/ε)) iterations. This coincides with the currently best-known iteration bounds for IPMs based on all existing kind of KFs.The algorithm based on the proposed KF has been tested. Extensive numerical simulations on test problems with different sizes have shown that this KF has promising results.展开更多
文摘This paper presents a stochastic modification of a limited memory BFGS method to solve bound-constrained global minimization problems with a differentiable cost function with no further smoothness. The approach is a stochastic descent method where the deterministic sequence, generated by a limited memory BFGS method, is replaced by a sequence of random variables. To enhance the performance of the proposed algorithm and make sure the perturbations lie within the feasible domain, we have developed a novel perturbation technique based on truncating a multivariate double exponential distribution to deal with bound-constrained problems;the theoretical study and the simulation of the developed truncated distribution are also presented. Theoretical results ensure that the proposed method converges almost surely to the global minimum. The performance of the algorithm is demonstrated through numerical experiments on some typical test functions as well as on some further engineering problems. The numerical comparisons with stochastic and meta-heuristic methods indicate that the suggested algorithm is promising.
文摘Interior-point methods(IPMs) for linear programming(LP) are generally based on the logarithmic barrier function. Peng et al.(J. Comput. Technol. 6: 61–80, 2001) were the first to propose non-logarithmic kernel functions(KFs) for solving IPMs. These KFs are strongly convex and smoothly coercive on their domains.Later, Bai et al.(SIAM J. Optim. 15(1): 101–128, 2004) introduced the first KF with a trigonometric barrier term. Since then, no new type of KFs were proposed until 2020, when Touil and Chikouche(Filomat. 34(12):3957–3969, 2020;Acta Math. Sin.(Engl. Ser.), 38(1): 44–67, 2022) introduced the first hyperbolic KFs for semidefinite program(ming(SD)P). They( establishe)d that the iteration complexities of algorithms based on their proposed KFs are O(n2/3log(n/ε) and O(n3/4log(n/ε)) for large-update methods, respectively. The aim of this work is to improve the complexity result for large-update method. In fact, we present a new parametric KF with a hyperbolic barrier term. By simple tools, we show that the worst-case iteration complexity of our algorithm for the large-update method is O(√n log n log(n/ε)) iterations. This coincides with the currently best-known iteration bounds for IPMs based on all existing kind of KFs.The algorithm based on the proposed KF has been tested. Extensive numerical simulations on test problems with different sizes have shown that this KF has promising results.
