Abstract. We present the Acyclic Bayesian Net Generator, a new approach to learn the structure of a Bayesian network using genetic algorithms. Due to the encoding mechanism, acycli...
In this paper we describe a methodology that includes the complementary use of simulated annealing and response surface methodology (RSM). The methodology was developed for analys...
In recent years there has been a flurry of works on learning Bayesian networks from data. One of the hard problems in this area is how to effectively learn the structure of a beli...
A constraint satisfaction problem, or CSP, can be reformulated as an integer linear programming problem. The reformulated problem can be solved via polynomial multiplication. If t...
Every day, people widely use information retrieval (IR) systems to answer their questions. We utilized the TREC 2007 complex, interactive question answering (ciQA) track to measur...