A Data Stream Outlier Detection Algorithm Based on Grid

The main aim of data stream outlier detection is to find the data stream outliers in rational time accurately. The existing outlier detection algorithms can find outliers in static data sets efficiently, but they are inapplicable for the dynamic data stream, and cannot find the abnormal data effectively. Due to the requirements of real-time detection, dynamic adjustment and the inapplicability of existing algorithms on data stream outlier detection, we propose a new data stream outlier detection algorithm, ODGrid, which can find the abnormal data in data stream in real time and adjust the detection results dynamically. According to the experiments on real datasets and synthetic datasets, ODGrid is superior to the existing data stream outlier detection algorithms, and it has good scalability to the dimensionality of data space.