摘要
支持向量机是通过求解对偶问题来解决原始问题的.针对线性决策函数f(x)=(w·x)+b,我们指出了其原有的逻辑系统中的错误,并通过严格的证明,对其理论基础作了改进.而且,对于阈值b,我们给出了一个新的简洁计算公式.
The basic Support Vector Machine for Classification solves the primal problem by solving the dual problem. Considering the linear decision function f(x)=(w·x)+b, an essential drawback in its logic system is pointed out and a strict theoretical foundation is established. Furthermore, for computing the threshold b, a new compact formula is proposed first time.
出处
《运筹学学报》
CSCD
北大核心
2004年第2期66-71,共6页
Operations Research Transactions
基金
This work is supported by the National Natural Science Foundation of China (No. 10371131).