IMPLEMENTATION OF CLUSTERING-BASED FEATURE SUBSET SELECTION ALGORITHM-FAST

Abstract

A FAST Algorithm produces the subset of most important features from the available set of features. Working of the FAST is done in two steps. In the primary step, the features are divided into clusters by using graphical method. In the secondary step, the most representative’s features are selected from each cluster. The feature selection algorithm is implemented from both point of views, among that one is the efficiency which is nothing but the time required to find subsets of the features and another is effectiveness which is related to the quality of the subset of features. When we apply the FAST algorithm on microarray data or any text data then it will not only give required subsets of features but also improves the performance. Feature section means to identify a required and most useful data and that gives the only required features from the databases