Global optimization based on noisy evaluations: An empirical study of two statistical approaches
Global optimization based on noisy evaluations: An empirical study of two statistical approaches
复制标题
基于噪声评估的全局优化：两种统计方法的实证研究
DOI：
10.1088/1742-6596/135/1/012100
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
E. Walter
中科院分区：
文献类型：
--
作者：
E. Vázquez;Julien Villemonteix;Maryan Sidorkiewicz;E. Walter
复杂的计算机代码的输出的优化往往是用少量的评估预算来实现的。已经开发并比较了专门用于此类问题的算法，例如预期改进算法（El）或全局优化信息方法（IAGO）。然而，这些比较的结果上的噪声评价结果的影响往往被忽视，尽管它经常出现在工业问题。在本文中，经验收敛速度El和IAGO进行比较时，加性噪声损坏的评估结果。IAGO似乎比El和El的各种修改更有效，旨在处理嘈杂的评价。关键词整体最佳化;电脑模拟;克里格法;高斯过程;杂讯评估。
The optimization of the output of complex computer codes has often to be achieved with a small budget of evaluations. Algorithms dedicated to such problems have been developed and compared, such as the Expected Improvement algorithm (El) or the Informational Approach to Global Optimization (IAGO). However, the influence of noisy evaluation results on the outcome of these comparisons has often been neglected, despite its frequent appearance in industrial problems. In this paper, empirical convergence rates for El and IAGO are compared when an additive noise corrupts the result of an evaluation. IAGO appears more efficient than El and various modifications of El designed to deal with noisy evaluations. Keywords. Global optimization; computer simulations; kriging; Gaussian process; noisy evaluations.