International Journal of Electrical and Computer Engineering (IJECE) Vol. 14, No. 5, October 2024, pp. 5984~5997 ISSN: 2088-8708, DOI: 10.11591/ijece.v14i5.pp5984-5997  5984 Journal homepage: http://ijece.iaescore.com Exploring the frontiers of trajectory outlier detection: an in-depth review and comparative analysis Sana Chakri 1 , Naoual Mouhni 2 , Faouzia Ennaama 1 1 LAMIGEP, Moroccan School of Engineering Sciences, Marrakech, Morocco 2 GL-ISI Team, Department of Informatics, Faculty of Science and Techniques, Moulay Ismail University, Meknes, Morocco Article Info ABSTRACT Article history: Received May 1, 2024 Revised Jul 10, 2024 Accepted Jul 17, 2024 This paper provides a review and comparative analysis of trajectory outlier detection methods. It presents the definition of outliers in trajectory data and the existing types to further examine the advanced approaches. Basic steps for detecting an outlier, which include data preprocessing, feature extraction, modeling, and similar, have been presented. Moreover, advanced methods such as autoencoders and the use of deep learning for outlier detection have been explored. In the end, this paper evaluates the techniques and compares them using common metrics, mainly focusing on the techniques based on autoencoders or deep learning. It covers applications in real life and practice along with any limitations, challenges, and perspective ideas for the future. Ultimately, it can be a useful resource for expanding the understanding of domain researchers and practitioners. Keywords: Autoencoder Cluster analysis Deep learning Outliers Pattern extraction This is an open access article under the CC BY-SA license. Corresponding Author: Sana Chakri LAMIGEP, Moroccan School of Engineering Sciences Marrakech 40000, Morocco Email: chakri.Sana@gmail.com 1. INTRODUCTION Detecting outliers in trajectories is a fundamental problem in many fields such as traffic monitoring, fleet management, and intrusion detection. The analysis of a trajectory as a sequence of points that describes the movement of an object in space and time helps identify abnormal behavior or rare and significant events. Detecting outliers in trajectories is critical to ensuring safety, optimizing resources, preventing accidents, and promoting informed decision-making. Trajectories represent the movement of objects in space and time, and identifying outliers on trajectories aims to detect abnormal behaviors or rare and significant events. Trajectory anomalies can be defined as typical behaviors or movements compared to normal conditions. These may include sudden changes in direction, unusual speeds, or deviations from expected spatial patterns. Figure 1 illustrates the case of expected trajectory outliers that can be extracted between two regions. Outliers are expected to be sub-trajectories that have some neighbors nearby, while normal trajectories have more neighbors nearby [1]. The initial part of this extensive research focuses on the significant issue of identifying outliers in trajectories, which is crucial in various fields such as traffic monitoring, fleet management, and intrusion detection [2]. By examining trajectories as sequences of points that represent an object's movement in both space and time [3], [4], it becomes feasible to detect abnormal behavior or rare, significant events. The ability to detect outliers within trajectories is essential for ensuring safety, optimizing resources, preventing accidents, and facilitating informed decision-making. Our main emphasis in this study is on the detection of outliers within trajectories, with a specific focus on recent advancements that encompass classical methods, machine learning algorithms, and deep learning-based techniques. We explore various definitions of