Answering queries in hybrid Bayesian networks using importance sampling

Fernández, Antonio; Rumí, Rafael; Salmerón, Antonio

doi:10.1016/j.dss.2012.03.007

Cited by 7 publications

(7 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…zeros in the probability tables. In scenarios of extreme probabilities, EW is known to be not so accurate [16], and therefore more sophisticated methods as the ones proposed in [4] for mixtures of truncated exponentials, are to be developed.…”

Section: Resultsmentioning

confidence: 99%

“…(2) is just the probability of evidence, which has to be estimated as well in order to have an answer to a query (recall thatθ 1 is just an estimator of the numerator). It was shown in [4] that numerator and denominator can be estimated using the same sample. To achieve this, instead of taking a sampling distribution defined on (a, b) it must be defined on the entire range of X i .…”

Section: Importance Samplingmentioning

confidence: 99%

“…(8), it determines the accuracy of the estimation. The rule is that the closer p * is to h, the lower the variance is [4].…”

Section: Importance Samplingmentioning

confidence: 99%

See 2 more Smart Citations

Parallel Importance Sampling in Conditional Linear Gaussian Networks

Salmerón

Ramos-López

Borchani

et al. 2015

Advances in Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

Abstract. In this paper we analyse the problem of probabilistic inference in CLG networks when evidence comes in streams. In such situations, fast and scalable algorithms, able to provide accurate responses in a short time are required. We consider the instantiation of variational inference and importance sampling, two well known tools for probabilistic inference, to the CLG case. The experimental results over synthetic networks show how a parallel version importance sampling, and more precisely evidence weighting, is a promising scheme, as it is accurate and scales up with respect to available computing resources.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Importance Samplingmentioning

confidence: 99%

See 1 more Smart Citation

Parallel Importance Sampling in Conditional Linear Gaussian Networks

Salmerón

Ramos-López

Borchani

et al. 2015

Advances in Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

show abstract

“…An approximate inference scheme was proposed by Rumí and Salmerón (2005), where the Penniless propagation algorithm (Cano et al, 2000) is adapted to MTEs. A more recent proposal for inference in hybrid BNs with MTEs (Fernández, Rumí, & Salmerón, 2012) is based on importance sampling with approximate pre-computation , where the sampling distributions are obtained following a variable elimination scheme (Zhang & Poole, 1996) and the distributions are represented using mixed trees (Moral, Rumí, & Salmerón, 2003). The instantiation of the Shenoy-Shafer architecture to MoPs was reported by Shenoy and West (2011a).…”

Section: Translation Proceduresmentioning

confidence: 99%

A Review of Inference Algorithms for Hybrid Bayesian Networks

Salmerón¹,

Rumí²,

Langseth³

et al. 2018

jair

Self Cite

View full text Add to dashboard Cite

Hybrid Bayesian networks have received an increasing attention during the last years. The difference with respect to standard Bayesian networks is that they can host discrete and continuous variables simultaneously, which extends the applicability of the Bayesian network framework in general. However, this extra feature also comes at a cost: inference in these types of models is computationally more challenging and the underlying models and updating procedures may not even support closed-form solutions. In this paper we provide an overview of the main trends and principled approaches for performing inference in hybrid Bayesian networks. The methods covered in the paper are organized and discussed according to their methodological basis. We consider how the methods have been extended and adapted to also include (hybrid) dynamic Bayesian networks, and we end with an overview of established software systems supporting inference in these types of models.

show abstract

“…We conjecture that this difference is caused by the presence of extreme probabilities in the season change model, which increases the variability of the sampling process carried out by IS making it more prone to error. A detailed discussion on the sensitivity of IS to extreme probabilities can be found in [4]. For longer sequences (50, 100 and 200 time steps), each experiment was replicated 100 times and the average metrics are presented in Table 3.…”

Section: Experimental Evaluationmentioning

confidence: 99%

MAP inference in dynamic hybrid Bayesian networks

et al. 2017

Self Cite

View full text Add to dashboard Cite

In this paper, we study the maximum a posteriori (MAP) problem in dynamic hybrid Bayesian networks. We are interested in finding the sequence of values of a class variable that maximizes the posterior probability given evidence. We propose an approximate solution based on transforming the MAP problem into a simpler belief update problem. The proposed solution constructs a set of auxiliary networks by grouping consecutive instantiations of the variable of interest, thus capturing some of the potential temporal dependences between these variables while ignoring others. Belief update is carried out independently in the auxiliary models, after which the results are combined, producing a configuration of to analyze the behavior of the approach. The algorithm has been implemented using Java 8 streams, and its scalability has been evaluated.

show abstract

Answering queries in hybrid Bayesian networks using importance sampling

Cited by 7 publications

References 46 publications

Parallel Importance Sampling in Conditional Linear Gaussian Networks

Parallel Importance Sampling in Conditional Linear Gaussian Networks

A Review of Inference Algorithms for Hybrid Bayesian Networks

MAP inference in dynamic hybrid Bayesian networks

Contact Info

Product

Resources

About