Wrappers for feature subset selection
Artificial Intelligence - Special issue on relevance
Hi-index | 0.00 |
The large-scale genome-wide SNP data being acquired from biomedical domains have offered resources to evaluate modern data mining techniques in applications to genetic studies. The purpose of this study is to extend our recently developed gene mining approach to extracting the relevant SNPs for alcoholism using sib-pair IBD profiles of pedigrees. Application to a publicly available large dataset of 100 simulated replicates for three American populations demonstrates that the proposed ensemble decision approach has successfully identified most of the simulated true loci, thus implicating that IBD statistic could be used as one of the informatics for mining the genetic underpins for complex human diseases.