Title :
Research on Classifying Performance of SVM with Modified Kernel Function in HCCR
Author :
Limin Sun ; Song, Yibin
Author_Institution :
Sch. of Comput. Sci. & Technol., Yantai Univ.
Abstract :
Support vector machines theoretically show very good performance for two-group classification problem, and the performance largely depends on the kernel function. However, there are no theories concerning how to choose good kernel functions based on practical using problem. In this paper, we tried to modify kernel both in data-dependent and margin-dependent way and applied the method to offline handwritten Chinese character recognition to investigate its classifying performance. Our experiment results show that the performance is improved with the proposed algorithm
Keywords :
handwritten character recognition; pattern classification; support vector machines; SVM modified kernel function; offline handwritten Chinese character recognition; support vector machines; two-group classification; Character recognition; Computer science; Electronic mail; Handwriting recognition; Kernel; Machine learning; Pattern recognition; Support vector machine classification; Support vector machines; Training data;
Conference_Titel :
Neural Networks and Brain, 2005. ICNN&B '05. International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-9422-4
DOI :
10.1109/ICNNB.2005.1614960
