Presentation | 1999/6/18 On a Clnange-of-Measture Algorithm for Rare-Event Estimation by Simulations Naoto Miyoshi, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | The importance sampling (IS) is well krnown as an effective technique to estimate rare-event probabilities via computer simulations. The main idea of IS is to change the probability distribution (change of measure) and make the event occurrence, which is originally rare, more frequent. In the IS setting, the choie of good distribution is crucial but the class of simulation models for which the optimal IS distribution is analytically calculated seems limited. In this note, we introduce some recent techniques to find (estimate) the optimal IS distribution through the simulation-based optimization approach. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Rare-event estimation / importance sampling / stochastic cournterpart method / regenerative sirmulation |
Paper # | IN99-25 |
Date of Issue |
Conference Information | |
Committee | IN |
---|---|
Conference Date | 1999/6/18(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Information Networks (IN) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | On a Clnange-of-Measture Algorithm for Rare-Event Estimation by Simulations |
Sub Title (in English) | |
Keyword(1) | Rare-event estimation |
Keyword(2) | importance sampling |
Keyword(3) | stochastic cournterpart method |
Keyword(4) | regenerative sirmulation |
1st Author's Name | Naoto Miyoshi |
1st Author's Affiliation | Department of Mathematical and Computing Sciences Tokyo Institute of Technology() |
Date | 1999/6/18 |
Paper # | IN99-25 |
Volume (vol) | vol.99 |
Number (no) | 127 |
Page | pp.pp.- |
#Pages | 7 |
Date of Issue |