Abstract
The assessment of the probability of a rare event with a naive Monte Carlo method is computationally intensive, so faster estimation or variance reduction methods are needed. We focus on one of these methods which is the interacting particle system (IPS) method. The method is not intrusive in the sense that the random Markov system under consideration is simulated with its original distribution, but selection steps are introduced that favor trajectories (particles) with high potential values. An unbiased estimator with reduced variance can then be proposed. The method requires to specify a set of potential functions. The choice of these functions is crucial because it determines the magnitude of the variance reduction. So far, little information was available on how to choose the potential functions. This paper provides the expressions of the optimal potential functions minimizing the asymptotic variance of the estimator of the IPS method and it proposes recommendations for the practical design of the potential functions.
References
[1] D. Aristoff, Analysis and optimization of weighted ensemble sampling, ESAIM Math. Model. Numer. Anal. 52 (2018), no. 4, 1219–1238. 10.1051/m2an/2017046Search in Google Scholar
[2] D. Aristoff and D. M. Zuckerman, Optimizing weighted ensemble sampling of steady states, Multiscale Model. Simul. 18 (2020), no. 2, 646–673. 10.1137/18M1212100Search in Google Scholar PubMed PubMed Central
[3] B. A. Berg, Introduction to multicanonical Monte Carlo simulations, Monte Carlo Methods (Toronto 1998), Fields Inst. Commun. 26, American Mathematical Society, Providence (2000), 1–24. 10.1090/fic/026/01Search in Google Scholar
[4] F. Cérou, P. Del Moral and A. Guyader, A nonasymptotic theorem for unnormalized Feynman–Kac particle models, Ann. Inst. Henri Poincaré Probab. Stat. 47 (2011), no. 3, 629–649. 10.1214/10-AIHP358Search in Google Scholar
[5] F. Cérou and A. Guyader, Adaptive multilevel splitting for rare event analysis, Stoch. Anal. Appl. 25 (2007), no. 2, 417–443. 10.1080/07362990601139628Search in Google Scholar
[6] F. Cérou, F. LeGland, P. Del Moral and P. Lezaud, Limit theorems for the multilevel splitting algorithm in the simulation of rare events, Proceedings of the Winter Simulation Conference, IEEE Press, Piscataway (2005), 682–691. Search in Google Scholar
[7] H. P. Chan and T. L. Lai, A general theory of particle filters in hidden Markov models and some applications, Ann. Statist. 41 (2013), no. 6, 2877–2904. 10.1214/13-AOS1172Search in Google Scholar
[8] N. Chopin, Central limit theorem for sequential Monte Carlo methods and its application to Bayesian inference, Ann. Statist. 32 (2004), no. 6, 2385–2411. 10.1214/009053604000000698Search in Google Scholar
[9] H. Chraibi, A. Dutfoy, T. Galtier and J. Garnier, On the optimal importance process for piecewise deterministic Markov process, ESAIM Probab. Stat. 23 (2019), 893–921. 10.1051/ps/2019015Search in Google Scholar
[10] P. Del Moral, Feynman–Kac Formulae. Genealogical and Interacting Particle Systems With Applications, Probab. Appl. (New York), Springer, New York, 2004. 10.1007/978-1-4684-9393-1Search in Google Scholar
[11] P. Del Moral, A. Doucet and A. Jasra, Sequential Monte Carlo samplers, J. R. Stat. Soc. Ser. B Stat. Methodol. 68 (2006), no. 3, 411–436. 10.1111/j.1467-9868.2006.00553.xSearch in Google Scholar
[12] P. Del Moral and J. Garnier, Genealogical particle analysis of rare events, Ann. Appl. Probab. 15 (2005), no. 4, 2496–2534. 10.1214/105051605000000566Search in Google Scholar
[13] R. Douc, E. Moulines and J. Olsson, Optimality of the auxiliary particle filter, Probab. Math. Statist. 29 (2009), no. 1, 1–28. Search in Google Scholar
[14] A. Doucet, N. de Freitas and N. Gordon, Sequential Monte Carlo Methods in Practice, Springer, New York, 2001. 10.1007/978-1-4757-3437-9Search in Google Scholar
[15] J. Garnier and P. Del Moral, Simulations of rare events in fiber optics by interacting particle systems, Opt. Commun. 267 (2006), 205–214. 10.1016/j.optcom.2006.05.066Search in Google Scholar
[16] M. Gerber, N. Chopin and N. Whiteley, Negative association, ordering and convergence of resampling methods, Ann. Statist. 47 (2019), no. 4, 2236–2260. 10.1214/18-AOS1746Search in Google Scholar
[17] P. Guarniero, A. M. Johansen and A. Lee, The iterated auxiliary particle filter, J. Amer. Statist. Assoc. 112 (2017), no. 520, 1636–1647. 10.1080/01621459.2016.1222291Search in Google Scholar
[18] D. Jacquemart and J. Morio, Tuning of adaptive interacting particle system for rare event probability estimation, Simul. Model. Practice Theory 66 (2016), 36–49. 10.1016/j.simpat.2016.02.004Search in Google Scholar
[19] H. R. Künsch, State space and hidden Markov models, Complex Stochastic Systems (Eindhoven 1999), Monogr. Statist. Appl. Probab. 87, Chapman & Hall/CRC, Boca Raton (2001), 109–173. 10.1201/9781420035988.ch3Search in Google Scholar
[20] H. R. Künsch, Recursive Monte Carlo filters: Algorithms and theoretical analysis, Ann. Statist. 33 (2005), no. 5, 1983–2021. 10.1214/009053605000000426Search in Google Scholar
[21] A. Lee and N. Whiteley, Variance estimation in the particle filter, Biometrika 105 (2018), no. 3, 609–625. 10.1093/biomet/asy028Search in Google Scholar
[22] J. Morio, D. Jacquemart, M. Balesdent and J. Marzat, Optimisation of interacting particle systems for rare event estimation, Comput. Statist. Data Anal. 66 (2013), 117–128. 10.1016/j.csda.2013.03.025Search in Google Scholar
[23] B. Peherstorfer, K. Willcox and M. Gunzburger, Survey of multifidelity methods in uncertainty propagation, inference, and optimization, SIAM Rev. 60 (2018), no. 3, 550–591. 10.1137/16M1082469Search in Google Scholar
[24] R. Y. Rubinstein, Simulation and the Monte Carlo Method, John Wiley & Sons, New York, 1981. 10.1002/9780470316511Search in Google Scholar
[25] J. Wouters and F. Bouchet, Rare event computation in deterministic chaotic systems using genealogical particle analysis, J. Phys. A 49 (2016), no. 37, Article ID 374002. 10.1088/1751-8113/49/37/374002Search in Google Scholar
© 2021 Walter de Gruyter GmbH, Berlin/Boston
