LHS in LHS: A new expansion strategy for Latin hypercube sampling in simulation design
Matteo Boschini, Davide Gerosa, Alessandro Crespi and Matteo Falcone
拉丁超级立方体采样(LHS)是模拟设计中突出的工具,在高维和计算昂贵的问题中具有各种应用。 LHS允许各种优化策略,最明显的是确保空间填充属性。 然而,LHS是一种单级算法,需要先验地了解目标样本量。 在这项工作中,我们在LHS中介绍了LHS,这是一种新的LHS扩展算法,可以在现有的LHS分布式集合中添加新样本,同时(大约)保留其属性。 总之,该算法识别远离初始集合的参数空间的区域,在这些区域中绘制新的LHS,然后将其与原始样本合并。 作为副产品,我们引入了一个新的度量,即LHS度,它量化给定设计与LHS分布的偏差。 我们的公开实现是通过 Python 包 expandLHS 分发的。
Latin Hypercube Sampling (LHS) is a prominent tool in simulation design, with a variety of applications in high-dimensional and computationally expensive problems. LHS allows for various optimization strategies, most notably to ensure space-filling properties. However, LHS is a single-stage algorithm that requires a priori knowledge of the targeted sample size. In this work, we present LHS in LHS, a new expansion algorithm for LHS that enables the addition of new samples to an existing LHS-distr...
