As historical data are typically unavailable for a start-up, risk assessment is always complex and challenging. Traditional methods incapable of capturing all facets this complexity; therefore, more sophisticated tools necessary. Using an expert-elicited Bayesian networks (BNs) methodology, paper aims to provide method combining diverse sources information, such as data, expert knowledge, the u...

Bayesian networks (BNs) have become an essential tool for reasoning under uncertainty in complex models. In particular, the subclass of Gaussian Bayesian networks (GBNs) can be used to model continuous variables with Gaussian distributions. Here we focus on the task of learning GBNs from data. Factorization of the multivariate Gaussian joint density according to a directed acyclic graph (DAG) p...

Bayesian network classifiers (BNC) have received considerable attention in machine learning field. Some special structure BNCs have been proposed and demonstrate promise performance. However, recent researches show that structure learning in BNs may lead to a non-negligible posterior problem, i.e, there might be many structures have similar posterior scores. In this paper, we propose a generali...

We examine the relationship between the Cooper-Herskovitz score of a Bayesian network and the conditional entropies of the nodes of the networks conditioned on the probability distributions of their parents. We show that minimizing the conditional entropy of each node of the BNS conditioned on its set of parents amounts to maximization of the CH score. The main result is a lower bound on the si...

The application of Bayesian networks (BNs) to cognitive assessment and intelligent tutoring systems poses new challenges for model construction. When cognitive task analyses suggest constructing a BN with several latent variables, empirical model criticism of the latent structure becomes both critical and complex. This paper introduces a methodology for criticizing models both globally (a BN in...

We redeene inference operations for multiply sectioned Bayesian networks (MS-BNs). When two adjacent subnets exchange belief, previous operations require repeated belief propagations within the receiving subnet. The new operations require such propagation only twice. We prove that the new operations do not compromise the coherence while improving the eeciency. A MSBN must be initialized before ...

This paper considers the accuracy of classification using Bayesian networks (BNs). It presents a method to localize network parts that are (i) in a given (rare) case responsible for a potential misclassification, or (ii) modeling errors that consistently cause misclassifications, even in common cases. We analyze how inaccuracies introduced by such network parts are propagated through a network ...

Correspondent inferences in attribution theory deal with assigning causes to behaviour based on true dispositions rather than situational factors. In this paper, we investigate how knowledge representation tools in Artificial Intelligence (AI), such as Bayesian networks (BNs), can help represent such situations and distinguish between the types of clues used in assessing the behaviour (disposit...

Learning Markov blanket (MB) structures has proven useful in performing feature selection, learning Bayesian networks (BNs), and discovering causal relationships. We present a formula for efficiently determining the number of MB structures given a target variable and a set of other variables. As expected, the number of MB structures grows exponentially. However, we show quantitatively that ther...

We propose statistical abduction as a rst-order logical framework for representing, inferring and learning probabilistic knowledge. It semantically integrates logical abduction with a parameterized distribution over abducibles. We show that statistical abduction combined with tabulated search provides an e cient algorithm for probability computation, a Viterbi-like algorithm for nding the most ...