Abstract | ||
---|---|---|
In this paper, we present a computationallyefficient method for inducing selectiveBayesian network classifiers. Our approachis to use information-theoretic metrics to efficientlyselect a subset of attributes fromwhich to learn the classifier. We explorethree conditional, information-theoretic metricsthat are extensions of metrics used extensivelyin decision tree learning, namely Quinlan's gain and gain ratio metrics and Mantaras's distance metric. We experimentallyshow that the... |
Year | Venue | Keywords |
---|---|---|
1996 | ICML | decision tree learning,distance metric |
Field | DocType | Citations |
Variable-order Bayesian network,Pattern recognition,Computer science,Wake-sleep algorithm,Influence diagram,Bayesian network,Artificial intelligence,Machine learning | Conference | 49 |
PageRank | References | Authors |
25.52 | 2 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Moninder Singh | 1 | 381 | 105.12 |
gregory provan | 2 | 503 | 120.02 |