A convex approach to optimum design of experiments with correlated observations

Optimal design of experiments for correlated processes is an increasingly relevant and active research topic. Until now only heuristic methods were available without a possibility to judge their quality. In this work we complement the virtual noise approach by a convex formulation and an equivalence theorem comparable to the uncorrelated case. Hence, it is now possible to provide an upper performance bound against which alternative design methods can be judged. We provide a comparison on some classical examples from the literature.

[1]  David C. Woods,et al.  Minimax Efficient Random Experimental Design Strategies With Application to Model-Robust Design for Prediction , 2020, Journal of the American Statistical Association.

[2]  L. Pronzato,et al.  Design of Experiments in Nonlinear Models: Asymptotic Normality, Optimality Criteria and Small-Sample Properties , 2013 .

[3]  Wolfgang Näther.  Effective observation of random fields , 1985 .

[4]  Christopher K. I. Williams,et al.  Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning) , 2005 .

[5]  W. Näther.  Optimum experimental designs , 1994 .

[6]  J. Sacks,et al.  Designs for Regression Problems with Correlated Errors III , 1966 .

[7]  Holger Dette,et al.  A new approach to optimal designs for correlated observations , 2015 .

[8]  Andrej Pázman,et al.  Criteria for optimal design of small-sample experiments with correlated observations , 2007, Kybernetika.

[9]  A. Zhigljavsky,et al.  A New Approach to Optimal Design for Linear Models With Correlated Observations , 2010 .

[10]  A. Pázman,et al.  Optimal design of experiments via linear programming , 2015, 1504.06226.

[11]  Holger Dette,et al.  Optimal designs in regression with correlated errors. , 2016, Annals of statistics.

[12]  Valerii Fedorov,et al.  16 Design of spatial experiments: Model fitting and prediction , 1996 .

[13]  Dariusz Ucinski,et al.  D-optimal sensor selection in the presence of correlated measurement noise , 2020 .

[14]  Michael Jackson,et al.  Optimal Design of Experiments , 1994 .

[15]  Peter Goos,et al.  Optimal Design of Experiments: A Case Study Approach , 2011 .

[16]  J. E. Kelley,et al.  The Cutting-Plane Method for Solving Convex Programs , 1960 .

[17]  Shalabh.  Design of Experiments: an Introduction based on Linear Models , 2012 .

[18]  D. Harville.  Matrix Algebra From a Statistician's Perspective , 1998 .

[19]  Andrej Pázman,et al.  Measures for designs in experiments with correlated errors , 2003 .

[20]  Jorge Mateu,et al.  Spatio-Temporal Design: Advances in Efficient Data Acquisition , 2012 .

[21]  A. Pázman,et al.  A New Interpretation of Design Measures , 1998 .

[22]  A. Zhigljavsky,et al.  The BLUE in continuous-time regression models with correlated errors , 2019, The Annals of Statistics.