Feature Selection using Functional Dependency (FSFD)

Abstract

While performing data mining we come across large number of attributes. The more the number of attributes more time is required to process it .Also it might lead to incorrect and unexpected outcomes ,as the number of attributes increases ,they might become irrelevant to each other. Before performing any data mining task, preprocessing needs to be done to remove the unwanted attributes. Feature selection is one of the dimensionality reduction techniques which can be used to complete the data mining task efficiently.  This paper throws light upon the importance of Feature selection technique in selecting the attributes of interest for the  successful completion of data mining task and puts forth a new algorithm for Feature selection FSFD .i.e. Feature Selection using Functional dependency.