关键词: 异常点检测, 局部敏感哈希, 随机游走, 数据挖掘

Abstract:

Outlier detection is one of the main research hotspots of data mining. There are many outlier detection methods, but the existing main outlier detection methods are inefficient in high dimensional data processing. In order to solve this problem, this paper proposes an efficient outlier detection algorithm. The algorithm combines the properties of locality sensitive Hashing and the random walks of the graph to identify outliers. Specifically, locality sensitive Hashing is used to efficiently process high-dimensional data, then the similarity is obtained by using the distance between data, and it is transformed into a transition probability of random walks. On this basis, the random walks technology is used to calculate the transfer probability between the data, in which the transfer probability between normal data will be higher and higher, and the probability of anomaly will be lower and lower, and then the anomaly data can be finally distinguished according to this property. The experimental results show that the pro-posed method can effectively detect outliers in the data and is generally better than other outlier detection algorithms.

Key words: outlier detection, locality sensitive Hashing, random walks, data mining