Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees
In this paper we introduce a new dynamic importance sampling propagation algorithm for Bayesian networks. Importance sampling is based on using an auxiliary sampling distribution from which a set of con gurations of the variables in the network is drawn, and the performance of the algorithm depends...
| Main Authors: | , |
|---|---|
| Format: | info:eu-repo/semantics/article |
| Language: | English |
| Published: |
2017
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10835/4893 https://doi.org/10.1016/j.ijar.2004.05.005 |
| _version_ | 1789406727332954112 |
|---|---|
| author | Moral, Serafín Salmerón Cerdán, Antonio |
| author_facet | Moral, Serafín Salmerón Cerdán, Antonio |
| author_sort | Moral, Serafín |
| collection | DSpace |
| description | In this paper we introduce a new dynamic importance sampling propagation algorithm for Bayesian networks. Importance sampling is based on using an auxiliary sampling distribution from which a set of con gurations of the variables in the network is drawn, and the performance of the algorithm depends on the variance of the weights associated with the simulated con gurations. The basic idea of dynamic importance sampling is to use the simulation of a con guration to modify the sampling
distribution in order to improve its quality and so reducing the variance of the future weights. The paper shows that this can be achieved with a low computational effort. The experiments carried out show that the nal results can be very good even in the case that the initial sampling distribution is far away from the optimum. |
| format | info:eu-repo/semantics/article |
| id | oai:repositorio.ual.es:10835-4893 |
| institution | Universidad de Cuenca |
| language | English |
| publishDate | 2017 |
| record_format | dspace |
| spelling | oai:repositorio.ual.es:10835-48932023-04-12T19:39:37Z Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees Moral, Serafín Salmerón Cerdán, Antonio Bayesian networks Probability propagation Approximate algorithms Importance sampling Probability trees In this paper we introduce a new dynamic importance sampling propagation algorithm for Bayesian networks. Importance sampling is based on using an auxiliary sampling distribution from which a set of con gurations of the variables in the network is drawn, and the performance of the algorithm depends on the variance of the weights associated with the simulated con gurations. The basic idea of dynamic importance sampling is to use the simulation of a con guration to modify the sampling distribution in order to improve its quality and so reducing the variance of the future weights. The paper shows that this can be achieved with a low computational effort. The experiments carried out show that the nal results can be very good even in the case that the initial sampling distribution is far away from the optimum. 2017-07-07T07:17:11Z 2017-07-07T07:17:11Z 2005 info:eu-repo/semantics/article http://hdl.handle.net/10835/4893 https://doi.org/10.1016/j.ijar.2004.05.005 en Attribution-NonCommercial-NoDerivatives 4.0 Internacional Attribution-NonCommercial-NoDerivatives 4.0 Internacional http://creativecommons.org/licenses/by-nc-nd/4.0/ info:eu-repo/semantics/openAccess |
| spellingShingle | Bayesian networks Probability propagation Approximate algorithms Importance sampling Probability trees Moral, Serafín Salmerón Cerdán, Antonio Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees |
| title | Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees |
| title_full | Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees |
| title_fullStr | Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees |
| title_full_unstemmed | Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees |
| title_short | Dynamic Importance Sampling in Bayesian Networks Based on Probability Trees |
| title_sort | dynamic importance sampling in bayesian networks based on probability trees |
| topic | Bayesian networks Probability propagation Approximate algorithms Importance sampling Probability trees |
| url | http://hdl.handle.net/10835/4893 https://doi.org/10.1016/j.ijar.2004.05.005 |
| work_keys_str_mv | AT moralserafin dynamicimportancesamplinginbayesiannetworksbasedonprobabilitytrees AT salmeroncerdanantonio dynamicimportancesamplinginbayesiannetworksbasedonprobabilitytrees |