Anomalous Trajectory Detection using Recurrent Neural Network

Abstract

Anomalous trajectory detection which plays an important role in taxi fraud detection and trajectory data preprocessing is a crucial task in trajectory mining fields. Traditional anomalous trajectory detection methods which utilize density and isolation approaches mainly focus on the differences of a new trajectory and the historical trajectory dataset. Although these methods can capture the particular characteristics of trajectories, they still suffer from the following two disadvantages. (1) These methods cannot capture the sequential information of the trajectory well. (2) These methods only concentrate on the given source and destination which may lead to data sparsity issues. To overcome those shortcomings, we propose a method called {\bf A}nomalous {\bf T}rajectory {\bf D}etection using {\bf R}ecurrent {\bf N}eural {\bf N}etwork (\textbf{ATD-RNN}) which characterizes the trajectory by learning the trajectory embedding. The trajectory embedding can capture the sequential information of the trajectory and depict the internal characteristics between anomalous and norm trajectory. To address the potential data sparsity problem, we enlarge the dataset between a source and a destination by taking the relevant trajectories into consideration. Extend experiments on real-world datasets validate the effectiveness of our method.

GeoTrackNet—A Maritime Anomaly Detector Using Probabilistic Neural Network Representation of AIS Tracks and A Contrario Detection

“…The latter group uses historical data to learn the implicit detection rules. Since no representative groundtruth data are available for maritime anomaly detection, learningbased anomaly detection schemes cannot apply supervised methods like in [5]- [7]. Unsupervised learning methods are then preferred [9]- [11], [16], [22], [23], [27].…”

“…For these reasons, anomaly detection methods used in other domains such as network traffic analysis or cybersecurity [3], [4] do not apply. We may also emphasise there are no representative groundtruth datasets for this task, hence, supervised learning strategies for anomaly detection as in [5]- [7] do not apply either.…”

Learning GPS Point Representations to Detect Anomalous Bus Trajectories

“…Next, a bag-of-roads method, similar to bag-of-words, generates vectors of the buses and vectors of the predefined routes, which are used by a KNN model to perform bus route classification. Song et al [31] introduce a binary classifier to detect spatial anomalies in taxi trajectories. For that, they first map trajectories into a regular grid map.…”

A Method for LSTM-Based Trajectory Modeling and Abnormal Trajectory Detection

“…Goh et al [9] proposed a novel rough neural network-based model named radial basis function network with dynamic decay adjustment (RBFNDDA) to learn information directly from a data set and group it in terms of prototypes, and then, a neighbourhood rough set-based procedure was applied to detect prototype outliers. Song et al [10] found the difficulty in learning the normal patterns and the problem of data sparsity and proposed anomalous trajectory detection using recurrent neural network (ATD-RNN) which learned the trajectory embedding to characterized the trajectory. Then, Cheng et al [11] used coordinate sequence and spatio-temporal sequence and proposed spatio-temporal recurrent neural network (ST-RNN) and added attention mechanism.…”

“…In practical application, we can find that the length of multiple trajectory data in the trajectory data set may not be the same, and there may be some problems such as missing data or different receiver working mechanism, resulting in data sparsity. When [10] researched on the same starting point trajectory modelling for trajectories with given departure and destination, they utilized some padding operations to align trajectories by using relevant trajectories. But this method needs enough prior knowledge of geography, increased the difficulty of data acquisition and data processing workload.…”