A quantile-based nested partition algorithm for black-box functions on a continuous domain
A quantile-based nested partition algorithm for black-box functions on a continuous domain
复制标题
DOI：
10.1109/wsc.2016.7822128
复制
发表时间：
2016-12
期刊：
影响因子：
--
通讯作者：
David D. Linz;Hao Huang;Z. Zabinsky
中科院分区：
文献类型：
--
作者：
David D. Linz;Hao Huang;Z. Zabinsky
仿真模型通常描述没有闭合形式的解析表示的复杂系统。本文提出了一种适合嵌套划分框架的连续区域上函数的算法，并使用分位数估计对区域进行排序和识别最有希望的区域。此外，我们应用最优计算预算分配(OCBA)方法来利用分位数估计量的正态性质来分配样本点。证明了对于满足Lipschitz条件的函数，该算法概率收敛到包含真正全局最优解的区域。文章最后给出了一些数值结果。
Simulation models commonly describe complex systems with no closed-form analytical representation. This paper proposes an algorithm for functions on continuous domains that fits into the nested partition framework and uses quantile estimation to rank regions and identify the most promising region. Additionally, we apply the optimal computational budget allocation (OCBA) method for allocating sample points using the normality property of quantile estimators. We prove that, for functions satisfying the Lipschitz condition, the algorithm converges in probability to a region that contains the true global optimum. The paper concludes with some numerical results.