Handbook of Statistics

Chapter 2

The Cross-Entropy Method for Estimation

Dirk P. Kroese¹, Reuven Y. Rubinstein² and Peter W. Glynn³,    ¹School of Mathematics and Physics, The University of Queensland, Brisbane 4072, Australia, ²Faculty of Industrial Engineering and Management, Technion, Haifa, Israel, ³Department of Management Science and Engineering, Stanford University, CA, USA, ^{[e-mail][e-mail][e-mail]}

Abstract

This chapter describes how difficult statistical estimation problems can often be solved efficiently by means of the cross-entropy (CE) method. The CE method can be viewed as an adaptive importance sampling procedure that uses the cross-entropy or Kullback–Leibler divergence as a measure of closeness between two ...