基于蚂蚁系统的数据流分类算法

数据流分类是数据挖掘中的重要问题,各种针对数据流分类的算法的提出,丰富了数据流挖掘的知识。而蚁群算法是模仿真实蚂蚁觅食行为而提出的一种具有高度创新性的启发元算法,随着其算法设计的不断改进,蚁群优化已成为组合优化领域最具潜力的算法之一。但是,很少有文章将两者联系在一起。本文提出了一种针对数据流分类的蚁群算法,很好地解决了数据流挖掘中的不确定性问题,给出了算法框架,并实现了分类生成、更新、合并和删除算法。在公共数据集上的验证证明算法具有较强的鲁棒性。

An ACO-Based Data Stream Classification Algorithm

Data stream classification is an important issue in data stream mining.Many algorithms for data stream classification have been proposed,which riches the knowledge of data stream mining.And ACO(ant colony optimization),which is enlightened by the ants' food-seeking behavior in the real world,is a very creative heuristic algorithm.With the development of ACO,it has been one of the most potential algorithms for the combination and optimization.This paper proposes a novel approach to clustering data streams wi...