A Tutorial on the Cross-Entropy Method

By P. Boer, Dirk P. Kroese, Shie Mannor, Reuven Y. Rubinstein. 2005

3,224 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research

How researchers see this paper

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

In Artificial Intelligence

In Management Science and Operations Research

Citations in the corpus

How to train your MAML (2018, methodology): “In this paper we use the cross-entropy (De Boer et al. (2005); Rubinstein (1999)) loss throughout.”
Benchmarking Model-Based Reinforcement Learning (2019, methodology): “Shooting Algorithms: RS is very effective on simple tasks such as InvertedPendulum, CartPole and Acrobot, but as task difficulty increases RS gradually gets surpassed by PETS-RS and PETS-CEM, which indicates that modelling uncertainty aware dynamics is crucial for the performance of shooting algorit”
Exploring Model-based Planning with Policy Networks (2019, methodology): “In PETS [5], the authors further use cross entropy method (CEM) [6, 1] to re-samples sequences near the best sequences from the last CEM iteration.”
HEAVY TAILS, IMPORTANCE SAMPLING AND CROSS–ENTROPY (2005, methodology): “In particular, this approach has proved fruitful for light tails where it most often leads to an exponential change of measure scheme.”
FACMAC: Factored Multi-Agent Centralised Policy Gradients (2020, methodology): “Specifically, CEM iteratively draws a batch of N random samples from a candidate distribution Dk, e.g., a Gaussian, at each iteration k.”
Adaptive methods for sequential importance sampling with application to state space models (2008, methodology): “Note that it is possible to approach this optimal sampling distribution by using the cross-entropy method ; see Rubinstein and Kroese (2004) and de Boer et al. (2005) and the references therein.”
A Review of Robot Learning for Manipulation: Challenges, Representations, and Algorithms (2019, background): “Actor-only policy search also includes gradient-free optimization methods [81, 257, 142, 80, 384], which are usable even when the policy is non-differentiable, and in some cases (e.”
Heterogeneous Face Attribute Estimation: A Deep Multi-Task Learning Approach (2017, methodology): “We handle nominal attributes in a classification scheme, and choose to use the cross-entropy loss [50]”
Edge Learning for B5G Networks With Distributed Signal Processing: Semantic Communication, Edge Computing, and Wireless Sensing (2022, neutral): “The objective function is defined by the cross-entropy (CE) [182] measuring the difference between p(X) and q(X).”
Safe end-to-end imitation learning for model predictive control (2018, methodology): “For all tasks, we provide positive terminal reward in the CEM algorithm if the task was completed successfully by the Bayesian network model or by the MPC-DDP controller. φT = { T, if success 0, otherwise (15) In order to encourage use of the learned model, we provide a positive immediate reward for”
Variational Inference with Tail-adaptive f-Divergence (2018, methodology): “…reverse KL divergence is used in expectation propagation (Minka, 2001; Opper & Winther, 2005), importance weighted auto-encoders (Burda et al., 2016), and the cross entropy method (De Boer et al., 2005); χ 2 -divergence is exploited for VI (e.g., Dieng et al., 2017), but is more extensively…”
Basis Function Adaptation in Temporal Difference Reinforcement Learning (2005, methodology): “This score can naturally serve as the score function for the CE method for optimization (Rubinstein, 1999; de-Boer et al., 2002).”
The Value Function Polytope in Reinforcement Learning (2019, methodology): “Gradient-free optimization methods have shown impressive performance over complex control tasks (De Boer et al., 2004; Salimans et al., 2017).”
The cross-entropy method for estimation (2013, methodology): “A tutorial on the CE method is given in de Boer et al. (2005).”
BASNet: Boundary-Aware Salient Object Detection (2019, methodology): “(ii) To obtain high confidence saliency map and clear boundary, we propose a hybrid loss that combines Binary Cross Entropy (BCE) [5], Structural SIMilarity (SSIM) [66] and IoU losses [42], which are expected to learn from ground truth information in pixel-, patch- and map- level, respectively.”
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].”