Inverse Problem for a Time-Series Valued Computer Simulator via Scalarization  

作　　者：Pritam Ranjan Mark Thomas Holger Teismann Sujay Mukhoti Pritam Ranjan;Mark Thomas;Holger Teismann;Sujay Mukhoti(OM & QT, Indian Institute of Management Indore, Indore, India;Department of Mathematics & Statistics, Acadia University, Wolfville, Canada)

机构地区：[1]OM & QT, Indian Institute of Management Indore, Indore, India [2]Department of Mathematics & Statistics, Acadia University, Wolfville, Canada

出　　处：《Open Journal of Statistics》2016年第3期528-544,共17页统计学期刊（英文）

摘　　要：For an expensive to evaluate computer simulator, even the estimate of the overall surface can be a challenging problem. In this paper, we focus on the estimation of the inverse solution, i.e., to find the set(s) of input combinations of the simulator that generates a pre-determined simulator output. Ranjan et al. [1] proposed an expected improvement criterion under a sequential design framework for the inverse problem with a scalar valued simulator. In this paper, we focus on the inverse problem for a time-series valued simulator. We have used a few simulated and two real examples for performance comparison.For an expensive to evaluate computer simulator, even the estimate of the overall surface can be a challenging problem. In this paper, we focus on the estimation of the inverse solution, i.e., to find the set(s) of input combinations of the simulator that generates a pre-determined simulator output. Ranjan et al. [1] proposed an expected improvement criterion under a sequential design framework for the inverse problem with a scalar valued simulator. In this paper, we focus on the inverse problem for a time-series valued simulator. We have used a few simulated and two real examples for performance comparison.

关 键 词：Calibration Computer Experiments Contour Estimation Gaussian Process Model Non-Stationary Process Sequential Design 

分 类 号：O17[理学—数学]