In the past decades, rapid growth of computer and database technologies has led to large-scale datasets. On other hand, data mining applications with high dimensional datasets that require speed accuracy are rapidly increasing. An important issue these is curse dimensionality, where number features much higher than patterns. One dimensionality reduction approaches feature selection can increase...

Most current feature selection techniques are focused on the incremental inclusion or exclusion of single individual features with respect to candidate subset(s). The use such approaches, where only inclusion/exclusion is considered, means that information as collaborative contribution correlation between may be lost. result final selected subset contain high levels inter-feature redundancy, as...