A Review of Modern Computational Algorithms for Bayesian Optimal Design

Bayesian experimental design is a fast growing area of research with many real-world applications. As computational power has increased over the years, so has the development of simulation-based design methods, which involve a number of algorithms, such as Markov chain Monte Carlo, sequential Monte Carlo and approximate Bayes methods, facilitating more complex design problems to be solved. The Bayesian framework provides a unified approach for incorporating prior information and/or uncertainties regarding the statistical model with a utility function which describes the experimental aims. In this paper, we provide a general overview on the concepts involved in Bayesian experimental design, and focus on describing some of the more commonly used Bayesian utility functions and methods for their estimation, as well as a number of algorithms that are used to search over the design space to find the Bayesian optimal design. We also discuss other computational strategies for further research in Bayesian optimal design.

[1]  I. Verdinelli.  Miscellanea. A note on Bayesian design for the normal linear model with unknown error variance , 2000 .

[2]  David Ríos Insua,et al.  Robust Bayesian analysis , 2000 .

[3]  Helga Wagner,et al.  Likelihood-free Simulation-based Optimal Design , 2013 .

[4]  X. Huan,et al.  GRADIENT-BASED STOCHASTIC OPTIMIZATION METHODS IN BAYESIAN EXPERIMENTAL DESIGN , 2012, 1212.2228.

[5]  James C. Spall,et al.  AN OVERVIEW OF THE SIMULTANEOUS PERTURBATION METHOD FOR EFFICIENT OPTIMIZATION , 1998 .

[6]  M. Degroot.  Optimal Statistical Decisions , 1970 .

[7]  W. J. Hill,et al.  A Joint Design Criterion for the Dual Problem of Model Discrimination and Parameter Estimation , 1968 .

[8]  Liam Paninski,et al.  Asymptotic Theory of Information-Theoretic Experimental Design , 2005, Neural Computation.

[9]  Peter A. J. Hilbers,et al.  Optimal experiment design for model selection in biochemical networks , 2014, BMC Systems Biology.

[10]  R. A. Leibler,et al.  On Information and Sufficiency , 1951 .

[11]  J. Geweke,et al.  Bayesian Inference in Econometric Models Using Monte Carlo Integration , 1989 .

[12]  W. J. Studden,et al.  Theory Of Optimal Experiments , 1972 .

[13]  H. Dette,et al.  Response-adaptive dose-finding under model uncertainty , 2011, 1107.5883.

[14]  David Lindley,et al.  Bayesian Statistics, a Review , 1987 .

[15]  Donald A. Berry,et al.  Group sequential clinical trials: a classical evaluation of Bayesian decision-theoretic designs , 1994 .

[16]  M. Wand,et al.  Explaining Variational Approximations , 2010 .

[17]  Michael P. H. Stumpf,et al.  Maximizing the Information Content of Experiments in Systems Biology , 2013, PLoS Comput. Biol..

[18]  J. Kiefer,et al.  Optimum Designs in Regression Problems , 1959 .

[19]  William J. Hill,et al.  Discrimination Among Mechanistic Models , 1967 .

[20]  N. Chopin.  A sequential particle filter method for static models , 2002 .

[21]  Samira Masoumi,et al.  Sequential Markov Chain Monte Carlo (MCMC) model discrimination , 2013 .

[22]  Teemu Pennanen,et al.  An Adaptive Importance Sampling Technique , 2006 .

[23]  Peter Müller,et al.  A Bayesian Decision-Theoretic Dose-Finding Trial , 2006, Decis. Anal..

[24]  Markus Hainy,et al.  Likelihood-Free Simulation-Based Optimal Design: An Introduction , 2014 .

[25]  Joseph B. Kadane,et al.  A Gridding Method for Bayesian Sequential Decision Problems , 2003 .

[26]  W. J. Studden,et al.  Bayesian Optimal Designs for Linear Regression Models , 1991 .

[27]  Emile H. L. Aarts,et al.  Simulated Annealing: Theory and Applications , 1987, Mathematics and Its Applications.

[28]  C. E. SHANNON,et al.  A mathematical theory of communication , 1948, MOCO.

[29]  Tuomas J. Lukka,et al.  Bayesian adaptive estimation: The next dimension , 2006 .

[30]  J. Bernardo.  Expected Information as Expected Utility , 1979 .

[31]  Jon Wakefield,et al.  An Expected Loss Approach to the Design of Dosage Regimens Via Sampling‐Based Methods , 1994 .

[32]  A. Pettitt,et al.  Approximate Bayesian computation using indirect inference , 2011 .

[33]  Chiara Tommasi,et al.  Bayesian optimum designs for discriminating between models with any distribution , 2010, Comput. Stat. Data Anal..

[34]  Y. Ho,et al.  Perturbation analysis and optimization of queueing networks , 1982, 1982 21st IEEE Conference on Decision and Control.

[35]  J. Kiefer.  General Equivalence Theory for Optimum Designs (Approximate Theory) , 1974 .

[36]  Robert J. Butera,et al.  Sequential Optimal Design of Neurophysiology Experiments , 2009, Neural Computation.

[37]  P. Müller,et al.  Optimal Bayesian Design by Inhomogeneous Markov Chain Simulation , 2004 .

[38]  J. Berger.  Statistical Decision Theory and Bayesian Analysis , 1988 .

[39]  Anthony N. Pettitt,et al.  Towards Bayesian experimental design for nonlinear models that require a large number of sampling times , 2014, Comput. Stat. Data Anal..

[40]  Peter Müller,et al.  Sequential stopping for high-throughput experiments , 2013, Biostatistics.

[41]  D Z D'Argenio,et al.  Incorporating prior parameter uncertainty in the design of sampling schedules for pharmacokinetic parameter estimation experiments. , 1990, Mathematical biosciences.

[42]  K. Chaloner.  Optimal Bayesian Experimental Design for Linear Models , 1984 .

[43]  Paul Gerhard Kinas,et al.  Bayesian fishery stock assessment and decision making using adaptive importance sampling , 1996 .

[44]  Dianpeng Wang,et al.  Sequential Bayesian Design for Estimation of EDp , 2009, 2009 2nd International Conference on Biomedical Engineering and Informatics.

[45]  H. Rue,et al.  Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations , 2009 .

[46]  Peter J. Diggle,et al.  Bayesian Geostatistical Design , 2006 .

[47]  Joseph B. Kadane,et al.  Optimal Experimental Design for Another's Analysis , 1993 .

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

[49]  Eamonn J. Keogh,et al.  Curse of Dimensionality , 2010, Encyclopedia of Machine Learning.

[50]  Friedrich Pukelsheim,et al.  Optimal weights for experimental designs on linearly independent support points , 1991 .

[51]  Stephen J. Wright,et al.  Numerical Optimization , 2018, Fundamental Statistical Inference.

[52]  C. Bielza,et al.  Decision Analysis by Augmented Probability Simulation , 1999 .

[53]  Xun Huan,et al.  Simulation-based optimal Bayesian experimental design for nonlinear systems , 2011, J. Comput. Phys..

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

[55]  J. L. Maryak,et al.  P Global Random Optimization by Simultaneous Perturbation Stochastic Approximation , 2005 .

[56]  Mark A. Pitt,et al.  Adaptive Design Optimization: A Mutual Information-Based Approach to Model Discrimination in Cognitive Science , 2010, Neural Computation.

[57]  E. Walter,et al.  Robust experiment design via stochastic approximation , 1985 .

[58]  Peter Müller,et al.  Optimal Design via Curve Fitting of Monte Carlo Experiments , 1995 .

[59]  James M. McGree,et al.  A pseudo-marginal sequential Monte Carlo algorithm for random effects models in Bayesian sequential design , 2016, Stat. Comput..

[60]  Drew Seils,et al.  Optimal design , 2007 .

[61]  C C Drovandi,et al.  Estimation of Parameters for Macroparasite Population Evolution Using Approximate Bayesian Computation , 2011, Biometrics.

[62]  W. J. Studden,et al.  Compromise designs in heteroscedastic linear models , 1992 .

[63]  H. Robbins.  A Stochastic Approximation Method , 1951 .

[64]  G. Gibson,et al.  Optimal Observation Times in Experimental Epidemic Processes , 2008, Biometrics.

[65]  J. Bernardo,et al.  Simulation-Based Optimal Design , 1999 .

[66]  Szu Hui Ng,et al.  Design of follow‐up experiments for improving model discrimination and parameter estimation , 2004 .

[67]  Christian P. Robert,et al.  Bayesian-Optimal Design via Interacting Particle Systems , 2006 .

[68]  K. Mengersen,et al.  Adaptive Bayesian compound designs for dose finding studies , 2012 .

[69]  Anthony N. Pettitt,et al.  Simulation-based fully Bayesian experimental design for mixed effects models , 2015, Comput. Stat. Data Anal..

[70]  H. Dette,et al.  A Generalization of $D$- and $D_1$-Optimal Designs in Polynomial Regression , 1990 .

[71]  Anthony C. Atkinson,et al.  DT-optimum designs for model discrimination and parameter estimation , 2008 .

[72]  T. Loredo.  Bayesian Adaptive Exploration , 2004, astro-ph/0409386.

[73]  Chiara Tommasi,et al.  Optimal designs for both model discrimination and parameter estimation , 2008 .

[74]  Anthony N Pettitt,et al.  Bayesian Experimental Design for Models with Intractable Likelihoods , 2013, Biometrics.

[75]  P. Bickel,et al.  Curse-of-dimensionality revisited: Collapse of the particle filter in very large scale systems , 2008, 0805.3034.

[76]  D. Lindley.  The Choice of Variables in Multiple Regression , 1968 .

[77]  Jason L. Loeppky,et al.  Batch sequential designs for computer experiments , 2010 .

[78]  David M. Borth,et al.  A Total Entropy Criterion for the Dual Problem of Model Discrimination and Parameter Estimation , 1975 .

[79]  Steven E. Rigdon,et al.  Model-Oriented Design of Experiments , 1997, Technometrics.

[80]  J L Palmer,et al.  Bayesian optimal design in population models for haematologic data. , 1998, Statistics in medicine.

[81]  J. López–Fidalgo,et al.  An optimal experimental design criterion for discriminating between non‐normal models , 2007 .

[82]  Peter Müller,et al.  Screening designs for drug development. , 2007, Biostatistics.

[83]  J. Zidek,et al.  Designing and integrating composite networks for monitoring multivariate gaussian pollution fields , 2000 .

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

[85]  P. Moral,et al.  Sequential Monte Carlo samplers , 2002, cond-mat/0212648.

[86]  A E Gelfand,et al.  Approaches for optimal sequential decision analysis in clinical trials. , 1998, Biometrics.

[87]  Peter D. H. Hill,et al.  A Review of Experimental Design Procedures for Regression Model Discrimination , 1978 .

[88]  K. Chaloner,et al.  The Equivalence of Constrained and Weighted Designs in Multiple Objective Design Problems , 1996 .

[89]  J Whitehead,et al.  Bayesian decision procedures based on logistic regression models for dose-finding studies. , 1998, Journal of biopharmaceutical statistics.

[90]  Heikki Haario,et al.  Simulation-Based Optimal Design Using a Response Variance Criterion , 2012 .

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

[92]  A. N. Pettitt,et al.  A sequential Monte Carlo approach to derive sampling times and windows for population pharmacokinetic studies , 2012, Journal of Pharmacokinetics and Pharmacodynamics.

[93]  John A. Nelder,et al.  A Simplex Method for Function Minimization , 1965, Comput. J..

[94]  J. Wolfowitz,et al.  On a Theorem of Hoel and Levine on Extrapolation Designs , 1965 .

[95]  Anthony N. Pettitt,et al.  Fully Bayesian Experimental Design for Pharmacokinetic Studies , 2015, Entropy.

[96]  D. Lindley.  On a Measure of the Information Provided by an Experiment , 1956 .

[97]  Peter Müller,et al.  Optimal sampling times in population pharmacokinetic studies , 2001 .

[98]  Kathryn Chaloner,et al.  Bayesian Experimental Design for Nonlinear Mixed‐Effects Models with Application to HIV Dynamics , 2004, Biometrics.

[99]  Raul Tempone,et al.  Fast estimation of expected information gains for Bayesian experimental designs based on Laplace approximations , 2013 .

[100]  D A Berry,et al.  One-sided sequential stopping boundaries for clinical trials: a decision-theoretic approach. , 1988, Biometrics.

[101]  Arnaud Doucet,et al.  SMC Samplers for Bayesian Optimal Nonlinear Design , 2006, 2006 IEEE Nonlinear Statistical Signal Processing Workshop.

[102]  Peter Müller,et al.  Bayesian randomized clinical trials: A decision‐theoretic sequential design , 2004 .

[103]  D. Balding,et al.  Approximate Bayesian computation in population genetics. , 2002, Genetics.

[104]  William F Rosenberger,et al.  Sequential designs for ordinal phase I clinical trials. , 2009, Biometrical journal. Biometrische Zeitschrift.

[105]  J. Kiefer.  Optimum Experimental Designs , 1959 .

[106]  J Whitehead,et al.  Bayesian decision procedures for dose determining experiments. , 1995, Statistics in medicine.

[107]  H. Rue,et al.  Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations , 2009 .

[108]  James M. McGree,et al.  Sequential Monte Carlo for Bayesian sequentially designed experiments for discrete data , 2013, Comput. Stat. Data Anal..

[109]  Anthony N. Pettitt,et al.  A Sequential Monte Carlo Algorithm to Incorporate Model Uncertainty in Bayesian Sequential Design , 2014 .

[110]  K. J. Ryan,et al.  Estimating Expected Information Gains for Experimental Designs With Application to the Random Fatigue-Limit Model , 2003 .

[111]  Anthony N. Pettitt,et al.  A sequential Monte Carlo approach to the sequential design for discriminating between rival continuous data models , 2012 .

[112]  Sang Joon Kim,et al.  A Mathematical Theory of Communication , 2006 .

[113]  S B Duffull,et al.  Optimal Design Criteria for Discrimination and Estimation in Nonlinear Models , 2009, Journal of biopharmaceutical statistics.