论文信息 - Research of Incremental Learning Algorithm Based on the Minimum Classification Error Criterion

Research of Incremental Learning Algorithm Based on the Minimum Classification Error Criterion

Incremental learning is widely used in artificial intelligence, pattern recognition and other fields. It is an effective method to solve the system, where samples are less in the beginning of training, but over time its performance reduces. In this paper, based on the analyses of support vector machine and the characteristics of incremental learning, we proposed incremental learning method which is based on the minimum classification error criterion. Moreover, the validity and feasibility of this algorithm is verified through experiments.

Xiusheng Duan | Ganlin Shan | Bo Wen

[1] Liu Zhi-jingb. Incremental learning algorithm based on SVM , 2007 .

[2] Wu Tie-jun. Fast incremental learning algorithm of linear support vector machine based on hull vectors , 2006 .

[3] Vladimir N. Vapnik,et al. The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[4] Cai Yu. Text categorization based on active learning support vector machines , 2009 .

[5] Zhang Bing,et al. A Fast Incremental Learning Algorithm for Support Vector Machine , 2005 .