Feature Selection for Ranking 数据挖掘课件.ppt

Feature Selection for Ranking  Cen fang ming 2007.10 ABSTRACT  The reality is that many feature selection methods used in classification are directly applied to ranking We argue that because of the striking differences between ranking and classification, it is better to develop different feature selection methods for ranking. In this paper we propose a new feature selection method Define the ranking accuracy in terms of a performance measure or a loss function as the importance of the feature and the correlation between the ranking results of two features as the similarity between them. We also demonstrate how to solve the optimization problem in an efficient way. We have tested the effectiveness of our feature selection method on two information retrieval datasets and with two ranking model 1. INTRODUCTION  Ranking SVM and RankNet First, feature selection can help enhance accuracy in many machine learning problems, which strongly indicates that feature selection is also necessary for ranking Second, feature selection can also help improve the efficiency of training. In information retrieval, especially in web search, usually the data size is very large and thus training of ranking models is computationally costly. There have been no methods of feature selection dedicatedly proposed for ranking. Most of the methods used in ranking were developed for classification. Feature selection methods in classification fall into three categorie The first category, which is named filter, feature selection is defined as a preprocessing step and can be independent from learning. that information gain (IG) and chi-square (CHI) are among the most effective methods of feature selection for classification. The second category referred to as wrapper utilizes the learning system as a black box to score subsets of features The third category called the embedded method performs feature selection within the process o