On the optimality and stability of exponential twisting in Monte Carlo estimation

Estimation of the large deviations probability p/sub n/=P(S/sub n/>or= gamma n) via importance sampling is considered, where S/sub n/ is a sum of n i.i.d. random variables. It has been previously shown that within the nonparametric candidate family of all i.i.d. (or, more generally, Markov) distributions, the optimized exponentially twisted distribution is the unique asymptotically optimal sampling distribution. As n to infinity , the sampling cost required to stabilize the normalized variance grows with strictly positive exponential rate for any suboptimal sampling distribution, while this sampling cost for the optimal exponentially twisted distribution is only O(n/sup 1/2/). Here, it is established that the optimality is actually much stronger. As n to infinity , this solution simultaneously stabilizes all error moments of both the sample mean and the sample variance estimators with sampling cost O(n/sup 1/2/). In addition, it is shown that the embedded parametric family of exponentially twisted distributions has a certain uniform asymptotic stability property. The technique is stable even if the optimal twisting parameter(s) cannot be precisely determined. >

[1]  R. R. Bahadur,et al.  On Deviations of the Sample Mean , 1960 .

[2]  U. Mosco On the continuity of the Young-Fenchel transform , 1971 .

[3]  D. Siegmund Importance Sampling in the Monte Carlo Study of Sequential Tests , 1976 .

[4]  R. Ellis,et al.  Entropy, large deviations, and statistical mechanics , 1985 .

[5]  P. Ney,et al.  MARKOV ADDITIVE PROCESSES II. LARGE DEVIATIONS , 1987 .

[6]  M. Woodroofe Nonlinear Renewal Theory in Sequential Analysis , 1987 .

[7]  P. Ney,et al.  Markov Additive Processes I. Eigenvalue Properties and Limit Theorems , 1987 .

[8]  Donald L. Iglehart,et al.  Importance sampling for stochastic simulations , 1989 .

[9]  Jean Walrand,et al.  A quick simulation method for excessive backlogs in networks of queues , 1989 .

[10]  James A. Bucklew,et al.  Large Deviation Techniques in Decision, Simulation, and Estimation , 1990 .

[11]  J.S. Sadowsky,et al.  On large deviations theory and asymptotically efficient Monte Carlo estimation , 1990, IEEE Trans. Inf. Theory.

[12]  P. Ney,et al.  Monte Carlo simulation and large deviations theory for uniformly recurrent Markov chains , 1990, Journal of Applied Probability.

[13]  J. S. Sadowsky,et al.  Direct-sequence spread-spectrum multiple-access communications with random signature sequences: A large deviations analysis , 1991, IEEE Trans. Inf. Theory.

[14]  J. Sadowsky Large deviations theory and efficient simulation of excessive backlogs in a GI/GI/m queue , 1991 .

[15]  Michael R. Frater,et al.  Optimally efficient estimation of the statistics of rare events in queueing networks , 1991 .

[16]  T. Lehtonen,et al.  SIMULATING LEVEL-CROSSING PROBABILITIES BY IMPORTANCE SAMPLING , 1992 .

[17]  Khaled Ben Letaief,et al.  Computing bit-error probabilities for avalanche photodiode receivers by large deviations theory , 1992, IEEE Trans. Inf. Theory.

[18]  Michael Devetsikiotis,et al.  An algorithmic approach to the optimization of importance sampling parameters in digital communication system simulation , 1993, IEEE Trans. Commun..