Optimization of computer simulation models with rare events

By Reuven Y. Rubinstein. 1997

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Mathematical Optimization

Citations in the corpus

Learning Latent Dynamics for Planning from Pixels (2018, methodology): “Planning algorithm We use the cross entropy method (CEM; Rubinstein, 1997; Chua et al., 2018) Initialize factorized belief over action sequences q ( a t : t + H ) ← Normal(0 , I ) . for optimization iteration i = 1 ..”
Modeling energy price dynamics: GARCH versus stochastic volatility (2016, methodology): “In this paper we follow Chan and Eisenstat (2015) and use an adaptive importance sampling method known as the cross-entropy method (Rubinstein, 1997; Rubinstein and Kroese, 2004) to compute the marginal likelihood.”
The cross-entropy method for estimation (2013, methodology): “The CE method was introduced by Rubinstein (1999, 2001), extending earlier work on variance minimization (Rubinstein, 1997).”
On the link between conscious function and general intelligence in humans and machines (2022, methodology): “Whilst such models could be queried with the agent’s current policy, in practice search algorithms like the cross-entropy method (Rubinstein, 1997; Hafner et al., 2019) or Monte Carlo tree search (MCTS) (Coulom, 2006; Schrittwieser et al.”
An adaptive heuristic cross-entropy algorithm for optimal design of water distribution systems (2007, methodology): “The cross-entropy (CE) method originates from an adaptive algorithm for rare events simulation based on variance minimization (Rubinstein 1997 ).”
A Tutorial on the Cross-Entropy Method (2005, methodology): “It was soon realized (Rubinstein, 1999, 2001) that a simple cross-entropy modification of Rubinstein (1997) could be used not only for estimating probabilities of rare events but for solving difficult COPs as well.”
Multifidelity importance sampling (2016, methodology): “Future work therefore includes an iterative approach, similar to the cross entropy method [7, 8], where the biasing distribution is repeatedly refined using samples of the biasing distribution of the previous iteration.”
Sample-Efficient Optimization in the Latent Space of Deep Generative Models via Weighted Retraining (2020, methodology): “The cross-entropy (CE) method iteratively retrains a generative model using a weighted training set, such that high-scoring points receive higher weights [57, 56, 12].”
V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning (2025, unknown): “…is repeated for several iterations before finally returning the mean of the sequence of Gaussians as the selected action trajectory. planning step using the Cross-Entropy Method (Rubinstein, 1997), and only execute the first action on the robot before re-planning, as in receding horizon control.”
Tactical supply chain planning under a carbon tax policy scheme: A case study (2015, unknown): “The Cross-Entropy (CE) method was first proposed by Rubinstein (1997) as a simulation method for estimating probabilities of rare events and was later adopted as an advanced optimization method to deal with both combinatorial and continuous optimization problems (Rubinstein, 1999).”
Temporal Difference Learning for Model Predictive Control (2022, methodology): “…found by iteratively ﬁtting parameters of a family of distributions, e.g., µ, σ for a multivariate Gaussian with diagonal covariance, to the space of actions over a ﬁnite horizon using the derivative-free Cross-Entropy Method (CEM; Rubinstein (1997)), and sample trajectories generated by a model.”
An introduction and survey of estimation of distribution algorithms (2011, methodology): “Developed independently of EDAs, the cross-entropy method (CE) [105] is probably the most closely related approach to EDAs and sees search much in the same way as EDAs.”