View Full Document

A Balanced Ensemble Approach to Weighting Classifiers for Text Classification



View the full content.
View Full Document
View Full Document

4 views

Unformatted text preview:

A Balanced Ensemble Approach to Weighting Classifiers for Text Classification Gabriel Pui Cheong Fung1 Jeffrey Xu Yu1 Haixun Wang2 David W Cheung3 Huan Liu4 1 The Chinese University of Hong Kong Hong Kong China pcfung yu se cuhk edu hk 2 IBM T J Watson Research Center New York USA haixun us ibm com 3 The University of Hong Kong Hong Kong China dcheung cs hku hk 4 Arizona State University Arizona USA hliu asu edu Abstract where d 1 indicates that d belongs to c and d 1 indicates that d does not belong to it f is a decision function Every classifier i has its own decision function fi If there are m different classifiers there will be m different decision functions The goal of constructing a binary classifier is to approximate the unknown true target function so that and are coincident as much as possible 17 In order to improve the effectiveness ensemble classifiers a k a classifier committee were proposed 1 3 5 6 7 8 9 15 16 17 18 19 An ensemble classifier is constructed by grouping a number of member classifiers If the decisions of the member classifiers are combined properly the ensemble is robust and effective There are two kinds of ensemble classifiers homogeneous and heterogeneous A homogeneous ensemble classifier contains m binary classifiers in which all classifiers are constructed by the same learning algorithm Bagging and boosting 19 are two common techniques 1 15 16 18 A heterogeneous ensemble classifier contains m binary classifiers in which all classifiers are constructed by different learning algorithms e g one SVM classifier and one kNN classifier are grouped together 19 The individual decisions of the classifiers in the ensemble are combined e g through stacking 19 1 if g 1 d 2 d m d 0 2 d 1 otherwise This paper studies the problem of constructing an effective heterogeneous ensemble classifier for text classification One major challenge of this problem is to formulate a good combination function which combines the decisions of the individual classifiers



Access the best Study Guides, Lecture Notes and Practice Exams

Loading Unlocking...
Login

Join to view A Balanced Ensemble Approach to Weighting Classifiers for Text Classification and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view A Balanced Ensemble Approach to Weighting Classifiers for Text Classification and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?