An Efficient Local Outlier Factor for Data Stream Processing: A Case Study

In the field of data mining and machine learning, outlier detection is considered a significant procedure for many applications, such as fraud detection in bank transactions and decision support systems. Data streams are a major player in the big data era. Currently, data streams are generated from various sources with huge amounts of data. This has led to difficulty when using older algorithms, which are designed for static data. The Local Outlier Factor (LOF) is one of these algorithms. The most challenging issue of the LOF is that it needs to preserve the whole dataset in computer memory. A new LOF that can deal with a data stream in limited memory is needed. This paper is a case study of several benchmark datasets for outlier detection that aim to increase the efficiency of the accuracy of local outlier detection in data streams.