基于纵横距离的单纯异常点检测算法及应用

Simple Outliers Detecting Algorithm Based on Vertical and Horizontal Distance and Its Application

首先讨论了异常点挖掘在数据挖掘过程中的重要性,产生异常点的原因,以及目前用于检测异常点的常用算法,指出了单纯应用距离法的局限性,提出了基于纵横距离的异常点检测算法,并给出了基于学生成绩检测的应用实例,该方法不需要进行大量的样本训练,在异常点检测方面有较好的效果.

This paper discusses the importance of outliers mining in the process of data mining,the cause of generating outliers and the common algorithms for detecting outliers.It points out the limitations of single distance algorithm,proposes an outliers detection algorithm based on vertical and horizontal distance and applies it to detect the students’score.This algorithm does not require a great number of training samples and has good performance on outliers’detection.