IncLOF:动态环境下局部异常的增量挖掘算法

异常检测是数据挖掘领域研究的最基本的问题之一，它在欺诈甄别、贷款审批、气象预报、客户分类等方面有广泛的应用，以前的异常检测算法只适应于静态环境，在数据更新时需要进行重新计算，在基于密度的局部异常检测算法LOF的基础上，提出一种在动态环境下局部异常挖掘的增量算法IncLOF，当数据库中的数据更新时，只对受到影响的点进行重新计算，这样可以大大提高异常的挖掘速度,实验表明，在动态环境下IncLOF的运行时间远远小于LOF的运行时间，并且用户定义的邻域中的最小对象个数与记录数之比越小，效果越明显.

IncLOF:An Incremental Algorithm for Mining Local Outliers in Dynamic Environment

Outlier detection is an important branch in the area of data mining It has been widely applied in fraud detection, loan approval, weather prediction and customer segmentation The former algorithms for outlier detection are only applied in static environment So recalculation is needed when updates occur In this paper, the first incremental outlier mining algorithm IncLOF is presented in dynamic environment based on LOF which is an algorithm for identifying density based local outliers IncLOF only recalculates a fraction of data affected by insertion or deletion, so it can greatly speed up outlier mining The results from a study on synthetic data sets demonstrate that the runtime of IncLOF is much less than LOF in dynamic environment, especially when the ratios of MinPts to database size are small