OFS Method for Selecting Active Features using Clustering Techniques
Thamarai Pugazhendhi Latchoumi, Jayakumar Loganathan, Latha Parthiban, Subbiah Janakiraman · 2016
Feature selection is the important method in data mining. In the traditional batch learning method, it is difficult to retrieve the selected features or data from the large set of dataset. The offline feature selection is difficult because it follows a priori. To overcome the offline selection features we are going to online feature selection by separating as a subset using classification and clustering techniques through that it will create a table and creates the attributes automatically. We cannot predict the range of the table accurately. So we are giving an average range, which can increase or decrease. It has the most relevant and correlated data it forms a subset and compares with the entropy and then it forms a tree construction by the subsets. Through the full input and partial input it produces the efficient and relevant result with any correlation. This will analyze the problem theoretically as well as practically using different data sets.