Improved cross-entropy method for estimation

摘要

The cross-entropy (CE) method is an adaptive importance sampling procedure that has been successfully applied to a diverse range of complicated simulation prob-lems. However, recent research has shown that in some high-dimensional settings, the likelihood ratio degeneracy prob-lem becomes severe and the importance sampling estima-tor obtained from the CE algorithm becomes unreliable. We consider a variation of the CE method whose perfor-mance does not deteriorate as the dimension of the prob-lem increases. We then illustrate the algorithm via a high-dimensional estimation problem in risk management.