Outlier Detection Algorithm Based on Quantitative Association

Outlier detection, aim of which is to discover the abnormal data objects in the data set, is an important research aspect in the data mining.This paper proposes a new method to detect outliers by discovering quantitative association rules between the data attributes.The outliers are defined as the small probability events which confidence less than the threshold in weak association rules.Experimental results show that the algorithm can effectively on outlier detection and has some application prospect.