POPE: post optimization posterior evaluation of likelihood free models

BMC Bioinformatics

Edward Meeds⁴  Michael Chiang²  Mary Lee¹  Olivier Cinquin²  John Lowengrub¹  Max Welling³ 

[1] Department of Mathematics, University of California, Irvine, USA

[2] School of Biological Sciences, University of California, Irvine, USA

[3] Donald Bren School of Informatics, University of California, Irvine, USA

[4] Informatics Institute, University of Amsterdam, Amsterdam, The Netherlands

关键词: Bayesian inference;    Simulation-based science;    Approximate Bayesian computation;   

Others  :  1229825
DOI  :  10.1186/s12859-015-0658-1

 received in 2015-02-09, accepted in 2015-07-02,  发布年份 2015

【 摘 要 】

Background

In many domains, scientists build complex simulators of natural phenomena that encode their hypotheses about the underlying processes. These simulators can be deterministic or stochastic, fast or slow, constrained or unconstrained, and so on. Optimizing the simulators with respect to a set of parameter values is common practice, resulting in a single parameter setting that minimizes an objective subject to constraints.

Results

We propose algorithms for post optimization posterior evaluation (POPE) of simulators. The algorithms compute and visualize all simulations that can generate results of the same or better quality than the optimum, subject to constraints. These optimization posteriors are desirable for a number of reasons among which are easy interpretability, automatic parameter sensitivity and correlation analysis, and posterior predictive analysis. Our algorithms are simple extensions to an existing simulation-based inference framework called approximate Bayesian computation. POPE is applied two biological simulators: a fast and stochastic simulator of stem-cell cycling and a slow and deterministic simulator of tumor growth patterns.

Conclusions

POPE allows the scientist to explore and understand the role that constraints, both on the input and the output, have on the optimization posterior. As a Bayesian inference procedure, POPE provides a rigorous framework for the analysis of the uncertainty of an optimal simulation parameter setting.

【 授权许可】

   
2015 Meeds et al.

【 图 表 】

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

Fig. 13.

Fig. 14.

Fig. 15.

Fig. 16.

Fig. 17.

Fig. 18.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

Fig. 13.

Fig. 14.

Fig. 15.

Fig. 16.

Fig. 17.

Fig. 18.

【 参考文献 】

• [1]Lizotte DJ. Practical Bayesian optimization. Phd thesis, University of Alberta; 2008.
• [2]Osborne MA, Garnett R, Roberts SJ. Gaussian processes for global optimization: 3rd International Conference on Learning and Intelligent Optimization (LION3); 2009, pp. 1–15.
• [3]Snoek J, Larochelle H, Adams RP. Practical Bayesian optimization of machine learning algorithms: Advances in Neural Information Processing Systems 25; 2012. arXiv:1206.2944.
• [4]Marjoram P, Molitor J, Plagnol V, Tavaré S. Markov chain Monte Carlo without likelihoods. Proc Nat Acad Sci. 2003; 100(26):15324-8.
• [5]Wilkinson R. Approximate Bayesian computation (ABC) gives exact results under the assumption of model error. Stat Appl Genet Mol Biol. 2013; 12(2):129-42.
• [6]Sisson SA, Fan Y. Likelihood-free Markov chain Monte Carlo. arXiv:1001.2058. 2010.
• [7]Jones DR, Schonlau M, Welch WJ. Efficient global optimization of expensive black-box functions. J Global Optimization. 1998; 13(4):455-92.
• [8]Couckuyt I, Gorissen D, DeTurck F, Dhaene T. Inverse surrogate modeling: output performance space sampling. In: 13th AIAA/ISSMO Multidisciplinary Analysis Optimization Conference. Texas: 2010.
• [9]Moral PD, Doucet A, O’Hagan A. Adaptive sequential Monte Carlo samplers: Technical report, University of Bordeaux; 2008.
• [10]Andrieu C, Roberts G. The pseudo-marginal approach for efficient Monte Carlo computations. Ann Stat. 2009; 37(2):697-725.
• [11]Bornn L, Pillai N, Smith A, Woodard D. The use of a single pseudo-sample in approximate Bayesian computation. ArXiv e-prints. 2014. arXiv:1404.6298v4.
• [12]Wood SN. Statistical inference for noisy nonlinear ecological dynamic systems. Nature. 2010; 466(7310):1102-4.
• [13]Turner BM, Sederberg PB. A generalized, likelihood-free method for posterior estimation. Psychon Bull Rev. 2014; 21(2):227-50.
• [14]Rasmussen CE. Gaussian Processes to speed up hybrid Monte Carlo for expensive Bayesian integrals. Bayesian Stat. 2003; 7:651-9.
• [15]Kennedy M, O’Hagan A. Bayesian calibration of computer models (with discussion). J R Stat Soc Ser B. 2001; 63:425-64.
• [16]Wilkinson R. Accelerating ABC methods using Gaussian processes. AISTATS. 2014. arXiv:1401.1436v2.
• [17]Meeds E, Welling M. GPS-ABC: Gaussian process surrogate approximate Bayesian computation. Uncertainty in AI. 2014. arXiv:1401.2838v1.
• [18]Beaumont MA, Cornuet JM, Marin JM, Robert CP. Adaptive approximate Bayesian computation. Biometrika. 2009; 96(4):983-90.
• [19]Silk D, Filippi S, Stumpf MPH. Optimizing threshold-schedules for sequential approximate Bayesian computation: applications to molecular systems. Stat Appl Genet Mol Biol. 2013; 12(5):603-18.
• [20]Bortot P, Coles S, Sisson S. Inference for stereological extremes. J Am Stat Assoc. 2007; 102:84-92.
• [21]Gelman A, Carlin JB, Stern HS, Rubin DB. Bayesian Data Analysis. Chapman and Hall/CRC, New York; 2004. Chap. 6
• [22]Lander AD, Gokoffski KK, Man FYM, Nie Q, Calof AL. Cell lineages and the logic of proliferative control. PLoS Biol. 2009; 7(1):608-19.
• [23]Itzkovitz S, Blat IC, Jacks T, Clevers H, van Oudenaarden A. Optimality in the development of intestinal crypts. Cell. 2012; 148(3):608-19.
• [24]Li L, Xie T. Stem cell niche: structure and function. Ann Rev Cell Dev Biol. 2005; 21:605-31.
• [25]Cinquin O. Purpose and regulation of stem cells: a systems-biology view from the Caenorhabditis elegans germ line. J Pathol. 2009; 217(2):608-19.
• [26]Pate KT, Stringari C, Sprowl-Tanio S, Wang K, TeSlaa T, Hoverter NP, et al. Wnt signaling directs a metabolic program of glycolysis and angiogenesis in colon cancer. EMBO J. 2014. doi:http://dx.doi.org/10.15252/embj.20148859810.15252/embj.201488598.