基于相反分类器的数据流分类方法

目前挖掘概念流动的数据流已经成为研究的热点。概念流动的数据流分类在预防信用卡欺诈，网络入侵发现等应用中具有重要的应用。本文定义了一种相反分类器来从错误中学习，提出了训练一个集合分类器来对具有概念流动的数据流进行分类的算法IWB。通过在合成数据集和benchmark上的实验，与Weighted Baggging算法比较，表明我们的算法具有更高的准确度，更快地收敛到新的目标概念的性能。

A Method for Classifying Data Stream Based on Reverse Classifier

Recently, mining data streams with concept drifts for actionable insights has become an important and challenging task for a wide range of applications including credit card fraud protection, network intrusion detection, etc. In this paper, we define the reverse classifier, which can help an algorithm to learn from error, and propose a IBW(improved weighted bagging) algorithm for classifying concept drift data streams using weighted ensemble classifiers. We evaluate IBW algorithm and Weighted Bagging algorithm on the STAGGER Concepts and synthetic data. The experiment results show that the proposed method have substantial advantage over Weighted Bagging approach in prediction accuracy, and it can converge to target concepts with high accuracy and speed.