Abstract
In real applications, a few unexpected examples unavoidably exist in the process of classification, not belonging to any known class. How to classify these unexpected ones is attracting more and more attention. However, traditional classification techniques can’t classify correctly unexpected instances, because the trained classifier has no knowledge about these. In this paper, we propose a novel entropy-based method to the problem. Finally, the experiments show that the proposed method outperforms previous work in the literature.
Keywords: Text Data, Severe Acute Respiratory Syndrome, Nominal Data, Positive Class, Negative Instance
Contributor Information
Qing Li, Email: itqli@cityu.edu.hk.
Ling Feng, Email: fengling@tsinghua.edu.cn.
Jian Pei, Email: jpei@cs.sfu.ca.
Sean X. Wang, Email: sean.wang@uvm.edu
Xiaofang Zhou, Email: zxf@itee.uq.edu.au.
Qiao-Ming Zhu, Email: qmzhu@suda.edu.cn.
Chaofeng Sha, Email: cfsha@fudan.edu.cn.
Zhen Xu, Email: xzhen@fudan.edu.cn.
Xiaoling Wang, Email: xlwang@sei.ecnu.edu.cn.
Aoying Zhou, Email: ayzhou@sei.ecnu.edu.cn.
References
- 1.Liu, B., Dai, Y., Li, X., Lee, W., Yu, S.: Building Text Classifiers Using Positive and Unlabeled Examples. In: IJCAI 2003 (2003)
- 2.Györfi L., Györfi Z., Vajda I. Bayesian decision with rejection. Problems of Control and Information Theory. 1978;8:445–452. [Google Scholar]
- 3.Devroye L., Györfi L., Lugosi G. A Probabilistic Theory of Pattern Recognition. Heidelberg: Springer; 1996. [Google Scholar]
- 4.Cover T., Thomas J. Elements of information theory. Hoboken: Wiley Interscience; 1991. [Google Scholar]
- 5.Li, X., Liu, B., Lee, W., Yu, S.: Text Classificaton by Labeling Words. In: AAAI 2004 (2004)
- 6.Li, X., Liu, B., Ng, S.: Learning to identify unexpected instances in the test set. In: IJCAI 2007 (2007)
- 7.Guo, Y., Greiner, R.: Optimistic active learning using mutual information. In: IJCAI 2007 (2007)