×

A multiple-try Metropolis-Hastings algorithm with tailored proposals. (English) Zbl 1505.62261

Summary: We present a new multiple-try Metropolis-Hastings algorithm designed to be especially beneficial when a tailored proposal distribution is available. The algorithm is based on a given acyclic graph \(\mathcal{G}\), where one of the nodes in \(\mathcal{G}\), \(k\) say, contains the current state of the Markov chain and the remaining nodes contain proposed states generated by applying the tailored proposal distribution. The Metropolis-Hastings algorithm alternates between two types of updates. The first type of update is using the tailored proposal distribution to generate new states for all nodes in \(\mathcal{G}\) except node \(k\). The second type of update is generating a new value for \(k\), thereby changing the value of the current state. We evaluate the effectiveness of the proposed scheme in two examples with previously defined target and proposal distributions.

MSC:

62-08 Computational methods for problems pertaining to statistics
65C05 Monte Carlo methods
65C40 Numerical analysis or methods applied to Markov chains

Software:

GSLIB
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] Abend K, Harley T, Kanal L (1965) Classification of binary random patterns. IEEE Trans Inf Theory 11:538-544 · Zbl 0129.11801 · doi:10.1109/TIT.1965.1053827
[2] Austad HM, Tjelmeland H (2017) Approximate computations for binary Markov random fields and their use in Bayesian models. Stat Comput 27:1271-1292 · Zbl 1505.62043 · doi:10.1007/s11222-016-9685-7
[3] Bédard M, Douc R, Moulines E (2012) Scaling analysis of multiple-try MCMC methods. Stoch Process Their Appl 122:758-786 · Zbl 1239.60075 · doi:10.1016/j.spa.2011.11.004
[4] Calderhead B (2014) A general construction for parallelizing Metropolis-Hastings algorithms. Proc Natl Acad Sci USA 111:17408-17413 · doi:10.1073/pnas.1408184111
[5] Casarin R, Craiu R, Leisen F (2013) Interacting multiple try algorithms with different proposal distributions. Stat Comput 23:1-16 · Zbl 1322.65003 · doi:10.1007/s11222-011-9301-9
[6] Chib S, Ramamurthy S (2010) Tailored randomized block MCMC methods with application to DSGE models. J Econ 155:19-38 · Zbl 1431.62603 · doi:10.1016/j.jeconom.2009.08.003
[7] Craiu RV, Lemieux C (2007) Acceleration of the multiple-try Metropolis algorithm using antithetic and stratified sampling. Stat Comput 17:109 · doi:10.1007/s11222-006-9009-4
[8] Cressie N, Davidson J (1998) Image analysis with partially ordered Markov models. Comput Stat Data Anal 29:1-26 · Zbl 1042.62611 · doi:10.1016/S0167-9473(98)00052-8
[9] Deutsch C, Journel A (1998) GSLIB: geostatistical software library, 2nd edn. Oxford University Press, Oxford
[10] Gamerman D, Lopes HF (2006) Markov chain Monte Carlo: stochastic simulation for Bayesian inference, 2nd edn. Chapman & Hall, London · Zbl 1137.62011
[11] Geman S, Geman D (1984) Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images. IEEE Trans Pattern Anal Mach Intell 6:721-741 · Zbl 0573.62030 · doi:10.1109/TPAMI.1984.4767596
[12] Gilks, WR; Bernardo, JM (ed.); Berger, JO (ed.); Dawid, AP (ed.); Smith, AFM (ed.), Derivative-free adaptive rejection sampling for Gibbs sampling, 641-649 (1992), Oxford
[13] Gilks WR, Richardson S, Spiegelhalter DJ (1996) Markov chain Monte Carlo in practice. Chapman & Hall, London · Zbl 0832.00018
[14] Green PJ (1995) Reversible jump MCMC computation and Bayesian model determination. Biometrika 82:711-732 · Zbl 0861.62023 · doi:10.1093/biomet/82.4.711
[15] Hammer PL, Rudeanu S (1968) Boolean methods in operation research and related areas. Springer, Berlin · Zbl 0155.28001 · doi:10.1007/978-3-642-85823-9
[16] Hastings WK (1970) Monte Carlo simulation methods using Markov chains and their applications. Biometrika 57:97-109 · Zbl 0219.65008 · doi:10.1093/biomet/57.1.97
[17] Journel A (1982) The indicator approach to estimation of spatial distributions. In: 17th APCOM symposium prooceedings. Society of Mining Engineers
[18] Liang F (2010) A double Metropolis-Hastings sampler for spatial models with intractable normalizing constants. J Stat Comput Simul 80:1007-1022 · Zbl 1233.62117 · doi:10.1080/00949650902882162
[19] Liang F, Liu C, Carroll R (2011) Advanced Markov chain Monte Carlo methods: learning from past samples. Wiley, New York · Zbl 1209.62009
[20] Liu JS, Liang FM, Wong WH (2000) The multiple-try method and local optimization in Metropolis sampling. J Am Stat Assoc 95:121-134 · Zbl 1072.65505 · doi:10.1080/01621459.2000.10473908
[21] Luo X, Tjelmeland H (2019) Prior specification for binary Markov mesh models. Stat Comput 29:367-389 · Zbl 1430.62185 · doi:10.1007/s11222-018-9813-7
[22] Mariethoz G, Caers J (2014) Multiple-point geostatistics: Stochastic modeling with training images, 1st edn. Wiley Blackwell, Chichester
[23] Martino L (2018) A review of multiple try MCMC algorithms for signal processing. Digit Signal Process 75:134-152 · doi:10.1016/j.dsp.2018.01.004
[24] Martino L, Read J (2013) On the flexibility of the design of multiple try Metropolis schemes. Comput Stat 28:2797-2823 · Zbl 1306.65099 · doi:10.1007/s00180-013-0429-2
[25] Martino L, Louzada F (2017) Issues in the multiple try Metropolis mixing. Comput Stat 32:239-252 · Zbl 1417.65015 · doi:10.1007/s00180-016-0643-9
[26] Martino L, Del Olmo VP, Read J (2012) A multi-point Metropolis scheme with generic weight functions. Stat Probab Lett 82:1445-1453 · Zbl 1315.65004 · doi:10.1016/j.spl.2012.04.008
[27] Martino L, Leisen F, Corander J (2014) On multiple try schemes and the particle Metropolis-Hastings algorithm. Tech. rep., ArXiv e-prints arXiv:1409.0051v1
[28] Metropolis N, Rosenbluth AW, Rosenbluth MN, Teller AH, Teller E (1953) Equation of state calculations by fast computing machines. J Chem Phys 21:1087-1092 · Zbl 1431.65006 · doi:10.1063/1.1699114
[29] Neal RN (2011) MCMC using ensemble of states for problems with fast and slow variables such as Gaussian process regression. Tech. rep., ArXiv e-prints arXiv:1101.0387v1
[30] Pandolfi S, Bartolucci F, Friel N (2014) A generalized multiple-try Metropolis version of the reversible jump algorithm. Comput Stat Data Anal 72:298-314 · Zbl 1506.62146 · doi:10.1016/j.csda.2013.10.007
[31] Pandolfi S, Bartolucci F, Friel N (2010) A generalization of the multiple-try Metropolis algorithm for Bayesian estimation and model selection. In: Teh YW, Titterington M (eds) Proceedings of the thirteenth international conference on artificial intelligence and statistics, vol. 9 of proceedings of machine learning research, PMLR, Sardinia, pp 581-588
[32] Qin ZS, Liu JS (2001) Multi-point Metropolis method with application to hybrid Monte Carlo. J Comput Phys 172:827-840 · Zbl 1158.65300 · doi:10.1006/jcph.2001.6860
[33] Riggan WB, Creason JP, Nelson WC, Manton KG, Woodbury MA, Stallard E, Pellom AC, Beaubier J (1987) U.S. cancer mortality rates and trends, 1950-1979. vol. IV (U.S. Goverment Printing Office, Washington, DC: Maps, U.S. Environmental Protection Agency)
[34] Robert CP, Casella G (1999) Monte Carlo statistical methods. Springer, Berlin · Zbl 0935.62005 · doi:10.1007/978-1-4757-3071-5
[35] Sherman M, Apanasovich TV, Carroll RJ (2006) On estimation in binary autologistic spatial models. J Stat Comput Simul 76:167-179 · Zbl 1088.62069 · doi:10.1080/00949650412331320873
[36] Stien M, Kolbjørnsen O (2011) Facies modeling using a Markov mesh model specification. Math Geosci 43:611-624 · Zbl 1219.86018 · doi:10.1007/s11004-011-9350-9
[37] Storvik G (2011) On the flexibility of Metropolis-Hastings acceptance probabilities in auxiliary variable proposal generation. Scand J Stat 38:342-358 · Zbl 1246.60097 · doi:10.1111/j.1467-9469.2010.00709.x
[38] Tjelmeland H, Hegstad BK (2001) Mode jumping proposals in MCMC. Scand J Stat 28:205-223 · Zbl 0972.65005 · doi:10.1111/1467-9469.00232
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.