[1]WANG Xiao-ming,WANG Shi-tong.Using average neighborhood margin with support vector machines[J].CAAI Transactions on Intelligent Systems,2010,5(4):313-319.
Copy
CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume:
5
Number of periods:
2010 4
Page number:
313-319
Column:
学术论文—人工智能基础
Public date:
2010-08-25
- Title:
-
Using average neighborhood margin with support vector machines
- Author(s):
-
WANG Xiao-ming; WANG Shi-tong
-
School of Information, Jiangnan University, Wuxi 214122, China
-
- Keywords:
-
supervised learning; support vector machine; neighborhood margin
- CLC:
-
TP181
- DOI:
-
-
- Abstract:
-
The support vector machine (SVM) is a learning method used successfully in the fields of machine learning and pattern recognition. However, traditional SVMs have not taken local information, or the environment of the data, into full consideration. To provide this context, an average neighborhood margin support vector machine (ANMSVM) was formulated by introducing the basic theories of neighborhood margins into SVMs. This method inherits the characteristics of traditional SVMs, yet also fully considers the local information surrounding the data, and thus shows better learning performance. Experimental results on artificial and real datasets indicated the effectiveness of an ANMSVM.