OPTIMAL DESIGN OF EXPERIMENTS WHEN FACTORS AFFECT DETECTION CAPABILITY

The configuration of a monitoring network affects the quality of information that is obtained. An important aspect is that some stations in a network may fail to provide data at all. In this work we study the important question of how to take account of detection capability in network configuration. Our approach is based on the theory of optimal design of experiments, extended to settings where detection must be considered. We illustrate our ideas by considering how to configure a seismic network that will provide precise estimates of the hypocenters of seismic events. An important consideration is that stations that are too far from the hypocenter may fail to detect the event. Our results provide a theoretical basis and practical guideline for configuring such a network.

[1]  Paul G. Richards,et al.  Quantitative Seismology: Theory and Methods , 1980 .

[2]  F. Pukelsheim.  Optimal Design of Experiments , 1993 .

[3]  David M. Steinberg,et al.  Optimal configuration of a seismographic network: A statistical approach , 1990, Bulletin of the Seismological Society of America.

[4]  K. Chaloner,et al.  Bayesian Experimental Design: A Review , 1995 .

[5]  L. White.  An extension of the General Equivalence Theorem to nonlinear models , 1973 .

[6]  Weng Kee Wong,et al.  Optimal design of experiments with anticipated pattern of missing observations. , 2004, Journal of theoretical biology.

[7]  H. Chernoff.  Locally Optimal Designs for Estimating Parameters , 1953 .

[8]  Weng Kee Wong,et al.  Bayesian Optimal Designs for a Quantal Dose-Response Study with Potentially Missing Observations , 2006, Journal of biopharmaceutical statistics.

[9]  D. Steinberg,et al.  Optimal seismic monitoring for event location with application to On Site Inspection of the Comprehensive Nuclear Test Ban Treaty , 2003 .

[10]  J. Kiefer,et al.  The Equivalence of Two Extremum Problems , 1960, Canadian Journal of Mathematics.

[11]  David M. Steinberg,et al.  Fast Computation of Designs Robust to Parameter Uncertainty for Nonlinear Settings , 2009, Technometrics.

[12]  David M. Steinberg,et al.  Configuring a seismographic network for optimal monitoring of fault lines and multiple sources , 1995, Bulletin of the Seismological Society of America.

[13]  Anthony C. Atkinson,et al.  Optimum Experimental Designs , 1992 .