Learning belief networks from data: an information theory based approach
CIKM '97 Proceedings of the sixth international conference on Information and knowledge management
Learning Bayesian networks from data: an information-theory based approach
Artificial Intelligence
Time and sample efficient discovery of Markov blankets and direct causal relations
Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining
Speculative Markov Blanket Discovery for Optimal Feature Selection
ICDM '05 Proceedings of the Fifth IEEE International Conference on Data Mining
Local learning algorithm for markov blanket discovery
AI'07 Proceedings of the 20th Australian joint conference on Advances in artificial intelligence
An efficient and scalable algorithm for local Bayesian network structure discovery
ECML PKDD'10 Proceedings of the 2010 European conference on Machine learning and knowledge discovery in databases: Part III
Learning Instance-Specific Predictive Models
The Journal of Machine Learning Research
Score-based methods for learning Markov boundaries by searching in constrained spaces
Data Mining and Knowledge Discovery
Learning the local Bayesian network structure around the ZNF217 oncogene in breast tumours
Computers in Biology and Medicine
| Hi-index | 0.00 |
Discovering the Markov blanket of a given variable can be viewed as a solution for optimal feature subset selection. Since 1996, several algorithms have been proposed to do local search of the Markov blanket, and they are proved to be much more efficient than the traditional approach where the whole Bayesian Network has to be learned first. In this paper, we compare those known published algorithms, including KS, GS, IAMB and its variants, PCMB, and one newly proposed called BFMB. We analyze the theoretical basis and practical values of each algorithm with the aim that it will help applicants to determine which ones to take in their specific scenarios.