Generalized Multiple Importance Sampling

Importance Sampling methods are broadly used to approximate posterior distributions or some of their moments. In its standard approach, samples are drawn from a single proposal distribution and weighted properly. However, since the performance depends on the mismatch between the targeted and the proposal distributions, several proposal densities are often employed for the generation of samples. Under this Multiple Importance Sampling (MIS) scenario, many works have addressed the selection or adaptation of the proposal distributions, interpreting the sampling and the weighting steps in different ways. In this paper, we establish a general framework for sampling and weighing procedures when more than one proposal are available. The most relevant MIS schemes in the literature are encompassed within the new framework, and, moreover novel valid schemes appear naturally. All the MIS schemes are compared and ranked in terms of the variance of the associated estimators. Finally, we provide illustrative examples which reveal that, even with a good choice of the proposal densities, a careful interpretation of the sampling and weighting procedures can make a significant difference in the performance of the method.

[1]  Jean-Michel Marin,et al.  Convergence of Adaptive Sampling Schemes , 2004 .

[2]  Jean-Marie Cornuet,et al.  Adaptive Multiple Importance Sampling , 2009, 0907.1254.

[3]  J. Marin,et al.  Population Monte Carlo , 2004 .

[4]  F. Liang Dynamically Weighted Importance Sampling in Monte Carlo Computation , 2002 .

[5]  O. Cappé,et al.  Population Monte Carlo , 2004 .

[6]  Jukka Corander,et al.  Layered adaptive importance sampling , 2015, Statistics and Computing.

[7]  A. Owen,et al.  Safe and Effective Importance Sampling , 2000 .

[8]  N. Gordon,et al.  Novel approach to nonlinear/non-Gaussian Bayesian state estimation , 1993 .

[9]  R. Douc,et al.  Minimum variance importance sampling via Population Monte Carlo , 2007 .

[10]  Jun S. Liu,et al.  Monte Carlo strategies in scientific computing , 2001 .

[11]  R. Douc,et al.  Convergence of Adaptive Sampling Schemes , 2007, 0708.0711.

[12]  Simon J. Godsill,et al.  On sequential Monte Carlo sampling methods for Bayesian filtering , 2000, Stat. Comput..

[13]  Mónica F. Bugallo,et al.  Heretical Multiple Importance Sampling , 2016, IEEE Signal Processing Letters.

[14]  Hera Y. He,et al.  Optimal mixture weights in multiple importance sampling , 2014, 1411.3954.

[15]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[16]  Luca Martino,et al.  Improving population Monte Carlo: Alternative weighting and resampling schemes , 2016, Signal Process..

[17]  P. McCullagh,et al.  A theory of statistical models for Monte Carlo integration , 2003 .

[18]  Leonidas J. Guibas,et al.  Optimally combining sampling techniques for Monte Carlo rendering , 1995, SIGGRAPH.

[19]  Zhiqiang Tan,et al.  On a Likelihood Approach for Monte Carlo Integration , 2004 .

[20]  Tim Hesterberg,et al.  Monte Carlo Strategies in Scientific Computing , 2002, Technometrics.

[21]  Eric Moulines,et al.  Comparison of resampling schemes for particle filtering , 2005, ISPA 2005. Proceedings of the 4th International Symposium on Image and Signal Processing and Analysis, 2005..

[22]  T. Hesterberg,et al.  Weighted Average Importance Sampling and Defensive Mixture Distributions , 1995 .

[23]  Christian P. Robert,et al.  Monte Carlo Statistical Methods , 2005, Springer Texts in Statistics.

[24]  Petar M. Djuric,et al.  Adaptive Importance Sampling: The past, the present, and the future , 2017, IEEE Signal Processing Magazine.

[25]  Heikki Haario,et al.  Adaptive proposal distribution for random walk Metropolis algorithm , 1999, Comput. Stat..

[26]  Mónica F. Bugallo,et al.  Adaptive importance sampling in signal processing , 2015, Digit. Signal Process..

[27]  Jukka Corander,et al.  An Adaptive Population Importance Sampler: Learning From Uncertainty , 2015, IEEE Transactions on Signal Processing.

[28]  Harald Niederreiter,et al.  Random number generation and Quasi-Monte Carlo methods , 1992, CBMS-NSF regional conference series in applied mathematics.

[29]  Mónica F. Bugallo,et al.  Efficient Multiple Importance Sampling Estimators , 2015, IEEE Signal Processing Letters.

[30]  Jean-Michel Marin,et al.  Adaptive importance sampling in general mixture classes , 2007, Stat. Comput..

[31]  Jukka Corander,et al.  A gradient adaptive population importance sampler , 2015, 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[32]  Jun S. Liu,et al.  Sequential Imputations and Bayesian Missing Data Problems , 1994 .

[33]  D. Owen Handbook of Mathematical Functions with Formulas , 1965 .

[34]  Philip Wagala Gwanyama The HM-GM-AM-QM Inequalities , 2004 .

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

[36]  Milton Abramowitz,et al.  Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Tables , 1964 .

[37]  H. Kahn,et al.  Methods of Reducing Sample Size in Monte Carlo Computations , 1953, Oper. Res..

[38]  Jukka Corander,et al.  An adaptive population importance sampler , 2014, 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[39]  H. Haario,et al.  An adaptive Metropolis algorithm , 2001 .