Interactive support vector machine learning algorithm and its application
Interactive support vector machine learning algorithm and its application
复制标题
DOI：
--
复制
发表时间：
1999
期刊：
影响因子：
--
通讯作者：
Li Yanda
中科院分区：
文献类型：
--
作者：
Li Yanda
交互式支持向量机（SVM）可以解决一些缺乏学习样本的监督学习问题。该算法基于SVM。在分类器的设计中引入了交互过程。根据现有的SVM分类器，可以主动选择一个“有用”的样本，从而得到更好的SVM分类器。与一般的SVM相比，它大大减少了样本数量，并且可能达到更高的泛化能力。在文本信息过滤上的应用就是该算法的成功范例。
Interactive support vector machine (SVM) can solve some supervised learning problems which are lack of learning samples. The algorithm bases on SVM. An interactive process is introduced in classifier's design. According to the prior SVM classifier, an “useful” sample can be select actively, then you can get the better SVM calssifier. Comparing with the general SVM, it greatly reduces the number of samples, and probably reaches higher generalization ability. The application on text information filtering shows a successful example using this algorithm.