The STUDIA UNIVERSITATIS BABEŞ-BOLYAI issue article summary

The summary of the selected article appears at the bottom of the page. In order to get back to the contents of the issue this article belongs to you have to access the link from the title. In order to see all the articles of the archive which have as author/co-author one of the authors mentioned below, you have to access the link from the author's name.

  
       
         
    STUDIA INFORMATICA - Issue no. Sp.Issue 1 / 2009  
         
  Article:   FEATURE SELECTION IN TEXT CATEGORIZATION USING ℓ1-REGULARIZED SVMS.

Authors:  ZSOLT MINIER. 		 
       
         
  Abstract:  Text categorization is an important task in the efficient handling of a large volume of documents. An important step in solving this task is the removal of certain features of the text that is not necessary for high precision classification. An interesting and well-founded method of feature selection are embedded methods that work directly on the hypothesis space of the machine learning algorithms used for classification. One such method is ℓ1-regularization that is used in conjunction with support vector machines. We study the effect of this method on precision of classiffying the 20 Newsgroups document corpus and compare it with the χ2 statistic feature selection method that is considered one of the best methods for feature selection in text categorization. Our findings show, that the ℓ1-regularization method performs about the same as the χ2statistic method.  
         
     
         
         
      Back to previous page