SIFStool is Social Inspired Feature Selection toolbox. It is a set of Matlab scripts and functions that can be used for feature subset selection problem. It is mainly based on the wrapper approach, where the selection criterion is an estimate of true missclassification error. The toolbox contains several search methods for the criterion minimization. One of these methods is the Simple Social Impact Theory based Optimizer (SSITO), which is inspired by a particular socio-psychological model of opinion formation in societies. Another implementation of different SITO methods was created by CSIO/CSIR and can be found here.



Previous page: PSGLab Project