Abstract: An efficient algorithm FCLBN for learning Bayesian network from small scale dataset was proposed. FCLBN uses the method of bootstrap to rcsample from the small scale dataset, and estimates the high confidence features of thesource small scale dataset from the Bayesian networks learned from the re-sampling small datasets. The high confidence features arc taken to guide the search of the best Baycsian network on the source dataset. After being evaluated on the standard benchmark dataset, FCLBN is applied to predict yeast protein localization. The result of the experiments indicafes that the FCLBN algorithm can learn relatively accurate network from small scale dataset.

Key words: Learning bayesian network,Small scale dataset,Features confidence