The following information was submitted:
Transactions: WSEAS TRANSACTIONS ON SYSTEMS
Transactions ID Number: 42-406
Full Name: Tao Jianwen
Position: Associate Professor
Age: ON
Sex: Male
Address: School of Information Engineer, Southern Yangtze University, Wuxi 214122, China
Country: CHINA
Tel:
Tel prefix:
Fax:
E-mail address: jianwen_tao@yahoo.com.cn
Other E-mails:
Title of the Paper: ¦Ñ-Margin Kernel Learning Machine with Magnetic Field Effect for both binary classification and novelty detection
Authors as they appear in the Paper:
Email addresses of all the authors:
Number of paper pages: 13
Abstract: A novel ¦Ñ-margin kernel learning machine (¦Ñ-MKLM) with magnetic field effect is proposed in allusion to pattern classification problem in this paper. In the Mercer induced feature space, ¦Ñ-MKLM can resolve pattern classification problems of both two-class and novelty detection. The basic idea is to find a optimal hyperplane with magnetic field effect such that the distance between one class (or normal samples) and the hyperplane is as small as possible due to the magnetic attractive effect, while at the same time the margin between the hyperplane and the other class (or abnormal samples ) is as large as possible due to magnetic repulsion, thus makeing binary patterns separated as much as possible. Moreover, an intensity factor q is also introduecd to improve the generalization performance of ¦Ñ-MKLM even more. To construct such a hyperplane with magnetic field effect, it is only needed for us to solve the corresponding convex quadratic programming problem with !
the off-the-shelf software pachages, such as LibSVM and PRTools etc. Exprimental results obtained from synthetic and real data show that the proposed algorithm is effective and competitive to other related algorithmss in such cases as two-class patterns classification and novelty detection problems respectively.
Keywords: magnetic field effect; pattern classification; novelty detection; support vector machine; kernel approach
EXTENSION of the file: .pdf
Special (Invited) Session:
Organizer of the Session:
How Did you learn about congress:
IP ADDRESS: 125.115.114.149
