Simply b & w of iheir rarity; theestimation of the statistic8 of buffer overflows in welldimensioned aueueine networks via direct skulation is extr&v costh. One technique that &be used to reduce this cost is importance samplf&, and it has been shown previously that Iarge deviations theory on be used in conjunction with importance sampling to minimize the required simulation time. In this paper,...

The Factored Policy-Gradient planner (FPG) (Buffet & Aberdeen 2006) was a successful competitor in the probabilistic track of the 2006 International Planning Competition (IPC). FPG is innovative because it scales to large planning domains through the use of Reinforcement Learning. It essentially performs a stochastic local search in policy space. FPG’s weakness is potentially long learning time...

Humans are able to accelerate their learning by selecting training materials that are the most informative and at the appropriate level of difficulty. We propose a framework for distributing deep learning in which one set of workers search for the most informative examples in parallel while a single worker updates the model on examples selected by importance sampling. This leads the model to up...

Importance sampling is a widely used variance reduction technique to compute sample quantiles such as value at risk. The variance of the weighted sample quantile estimator is usually a difficult quantity to compute. In this paper we present the exact convergence rate and asymptotic distributions of the bootstrap variance estimators for quantiles ofweighted empirical distributions. Under regular...

In this paper we analyze a particular issue of estimation, namely the estimation of the expected value of an unknown function for a given distribution, with the samples drawn from other distributions. A motivation of this problem comes from machine learning. In reinforcement learning, an intelligent agent that learns to make decisions in an unknown environment encounters the problem of judging ...

We propose simulation based estimation for discrete sequential move games of perfect information which relies on the simulated moments and importance sampling. We use importance sampling techniques not only to reduce computational burden and simulation error, but also to overcome non-smoothness problems. The model is identified with only weak scale and location normalizations, monte Carlo evide...

Stochastic Differential Equations (SDEs) are used as statistical models in many disciplines. However, intractable likelihood functions for SDEs make inference challenging, and we need to resort to simulation-based techniques to estimate and maximize the likelihood function. While importance sampling methods have allowed for the accurate evaluation of likelihoods at fixed parameter values, there...

Over the past several years importance sampling in conjunction with regenerative simulation has been presented as a promising method for estimating reliability measures in highly dependable Markovian systems. Existing methods fail to provide benefits over crude Monte Carlo for the analysis of systems that contain significant component redundancies. This paper presents refined importance samplin...

Two sequential methods are described for sampling constrained binary sequences from partial solutions. The backward method computes elimination ideals over finite fields and constructs partial solutions that extend. The forward method uses numerical global optimization to determine which partial solutions extend. The methods are applied to restricted orderings, binary dynamics, and random graphs.

We study a family of importance sampling estimators for the problem of computing the probability of level crossing when the crossing level is large, or when the intensity of the noise is small. We give general results concerning centered gaussian processes with drift and develop a method which allows to compute explicitly the asymptotics of the second order moment, with a special mention for th...