基于核心向量机的多任务概念漂移数据快速分类

The core vector machine-based rapid classification of multi-task concept drift dataset

通过协同求解多个概念漂移问题并充分挖掘相关概念漂移问题中蕴含的有效信息,共享矢量链支持向量机(shared vector chain supported vector machines,SVC-SVM)在面向多任务概念漂移分类时表现出良好性能。然而实际应用中的概念漂移问题通常有较大的数据容量,较高的计算代价限制了SVC-SVM方法的推广能力。针对这个弱点,借鉴核心向量机的近线性时间复杂度的优势,提出了适于多任务概念漂移大规模数据的共享矢量链核心向量机(shared vector chain core vector machines,SVC-CVM)。SVC-CVM具有渐近线性时间复杂度的算法特点,同时又继承了SVC-SVM方法协同求解多个概念漂移问题带来的良好性能,实验验证了该方法在多任务概念漂移大规模数据集上的有效性和快速性。

The shared vector chain-supported vector machine(SVC-SVM)can solve multiple concept drift problems as well as related problems,and it shows attractive performance in multi-task concept drift classification.However,in many practical scenarios,the concept drift dataset is usually large,and its high computational cost severely limits the generalization ability of the SVC-SVM.To overcome this shortcoming,a novel classifier termed shared vector chaincore vector machine(SVC-CVM)is proposed for large scale multi-task concept drift dataset,considering the asymptotic linear time complexity of the core vector machines.This classifier has the merit of asymptotic time complexity and inherits the good performance of SVC-SVM in solving multi-task concept drift problems.Furthermore,the effectiveness and rapidness of the proposed method is experimentally confirmed on large-scale multi-task concept drift datasets.