×

Learning Bayesian networks by ant colony optimisation: Searching in two different spaces. (English) Zbl 1036.68108

Summary: The most common way of automatically learning Bayesian networks from data is the combination of a scoring metric, the evaluation of the fitness of any given candidate network to the data base, and a search procedure to explore the search space. Usually, the search is carried out by greedy hill-climbing algorithms, although other techniques such as genetic algorithms, have also been used.
A recent metaheuristic, Ant Colony Optimisation (ACO), has been successfully applied to solve a great variety of problems, being remarkable the performance achieved in those problems related to path (permutation) searchin in graphs, such as the traveling salesman problem. In two previous works, the authors have approached the problem of learning Bayesian networks by means of the search+score methodology using ACO as the search engine.
As in these articles the search was performed in different search spaces, in the space of orderings and in the space of directed acyclic graphs. In this paper we compare both approaches by analysing the results obtained and the differences in the design and implementation of both algorithms.

MSC:

68T20 Problem solving in the context of artificial intelligence (heuristics, search strategies, etc.)
68T05 Learning and adaptive systems in artificial intelligence
PDFBibTeX XMLCite
Full Text: EuDML