Unsupervised learning trajectory anomaly detection algorithm based on deep representation

Zhongqiu Wang,Haoran Pei,Guan Yuan,Xiao Liu,Yanmei Zhang

doi:10.1177/1550147720971504

Abstract

Without ground-truth data, trajectory anomaly detection is a hard work and the result lacks of interpretability. Moreover, in most current methods, trajectories are represented by geometric features or their low-dimensional linear combination, and some hidden features and high-dimensional combined features cannot be found efficiently. Meanwhile, traditional methods still cannot get rid of the limitation of space attributes. Therefore, a novel trajectory anomaly detection algorithm is present in this article. Unsupervised learning mechanism is used to overcome nonground-truth problem and deep representation method is used to represent trajectories in a comprehensive way. First, each trajectory is partitioned into segments according to its open angles, then the shallow features at each point of a segment are extracted and. In this way, each segment is represented as a feature sequence. Second, shallow features are integrated into auto-encoder-based deep feature fusion model, and the fusion feature sequences can be extracted. Third, these fused feature sequences are grouped into different clusters using a unsupervised clustering algorithm, and then segments which quite differ from others are detected as anomalies. Finally, comprehensive experiments are conducted on both synthetic and real data sets, which demonstrate the efficiency of our work.

Highlights

With the continuous development and application of positioning equipment, such as global position system (GPS), Wi-Fi as well as wireless sensor network (WSN), a large volume of moving objects can be tracked with their trajectory data stored in the database
Compared with algorithms based on distance and density, the deep feature fusion model based on autoencoder is proposed in this article, which transforms the geometric feature representation of trajectory into sequence pattern, represents trajectory by feature sequence, and desalinates the spatial attribute of trajectory itself
In this article, labeled artificial trajectory data set is used first to verify the unsupervised trajectory anomaly detection algorithm based on deep representation; we compare the detected anomaly with the label in the data set

Summary

Introduction

With the continuous development and application of positioning equipment, such as global position system (GPS), Wi-Fi as well as wireless sensor network (WSN), a large volume of moving objects can be tracked with their trajectory data stored in the database. Compared with algorithms based on distance and density, the deep feature fusion model based on autoencoder is proposed in this article, which transforms the geometric feature representation of trajectory into sequence pattern, represents trajectory by feature sequence, and desalinates the spatial attribute of trajectory itself. In order to obtain the trajectory segments of different motion modes, the density-based spatial clustering of applications with noise (DBSCAN) algorithm is used to cluster by measuring the cosine distance between fusion feature sequences. The lim- ‘‘Feature extraction and transformation,’’ the deep feaitation of using spatial distance to judge the ture fusion model based on auto-encoder is proposed anomaly is fundamentally solved using the way to describe how to extract and transform the features of feature sequence to represent the trajectory from the divided trajectory segments. In section ‘‘Conclusion,’’ the full text is summarized, and the future research direction is pointed out

Related work

Connectivity

Experimental results analysis

Conclusion