duoshili

  考虑到传统的多示例多标签框架忽略了标签相关性，或者是将标签相关性作为全局概念知识加以利用，我们将局部标签相关性引入到MIML一SVM多示例多标签学习框架中来，得到了一种新的多示例多标签学习算法MIML LC SVM(Taking into account the traditional multi-sample multi-label framework ignores the label correlation or correlation between the label to be used as a global concept of knowledge, we will bow} local label correlation MIML into a multi-instance multi-label SVM learning framework in the past, get a new multi-instance multi-label learning algorithm MIML LC SVM)