摘要
提出一种基于遗传算法的个性化特征选择方法。该方法采用真伪两类样本之间的边缘间隔作为遗传算法的适应度估计函数,在相同特征初始集基础上对不同人提取不同的(即个性化)特征子集。实验证明该方法不但能有效地降低特征空间维数,而且使分类准确率得到显著提高。
A method of personal features selection based on genetic algorithm to improve the performance of the on-line signature verification systen is introduced. A new edge distance between the two classes is used as fitness evaluation function in GA. The personal sub features can be selected from the original feature group. Experimental results show that the method can not only reduce the dimension of the feature space but also advance the performance of the verification.
出处
《计算机应用与软件》
CSCD
北大核心
2007年第8期159-161,174,共4页
Computer Applications and Software