A note on imprecise Monte Carlo over credal sets via importance sampling

Troffaes, Matthias C.M.

Abstract

This brief paper is an exploratory investigation of how we can apply sensitivity analysis over importance sampling weights in order to obtain sampling estimates of lower previsions described by a parametric family of distributions. We demonstrate our results on the imprecise Dirichlet model, where we can compare with the analytically exact solution. We discuss the computational limitations of the approach, and propose a simple iterative importance sampling method in order to overcome these limitations. We find that the proposed method works pretty well, at least in the example studied, and we discuss some further possible extensions.

Citation

Troffaes, M. C. (2017, July). A note on imprecise Monte Carlo over credal sets via importance sampling. Presented at The Tenth International Symposium on Imprecise Probability: Theories and Applications (ISIPTA ’17), Lugano, Switzerland

• Presentation Conference Type: Conference Paper (published)
• Conference Name: The Tenth International Symposium on Imprecise Probability: Theories and Applications (ISIPTA ’17)
• Start Date: Jul 10, 2017
• End Date: Jul 14, 2017
• Acceptance Date: Apr 17, 2017
• Online Publication Date: Jun 20, 2017
• Publication Date: Jun 20, 2017
• Deposit Date: Mar 9, 2017
• Publicly Available Date: May 15, 2017
• Publisher: PMLR
• Pages: 325-332
• Series Title: Proceedings of Machine Learning Research
• Series Number: 62
• Series ISSN: 1938-7228
• Book Title: Proceedings of the Tenth International Symposium on Imprecise Probability : Theories and Applications, 10-14 July 2017, Lugano (Switzerland).
• Public URL: https://durham-repository.worktribe.com/output/1147530
• Publisher URL: http://proceedings.mlr.press/v62/troffaes17a.html