Group Abnormal Behavior Detection Based on Fuzzy Clustering

For group abnormal behavior analysis, previous methods are easy to be affected by the normal fluctuation characteristics of target or cause “missing” alarm. To overcome these problems, this paper proposes a group abnormal behavior detection method based on fuzzy clustering strategy, to improve the accuracy of abnormal trip behavior measurement and further mine the abnormal targets in the group. Experimental results demonstrate the efficiency of the proposed method on real-world datasets.