Unsupervised learning trajectory anomaly detection algorithm based on deep representation

Wang, Zhong-Qiu; Yuan, Guan; Pei, Haoran; Zhang, Yanmei; Liu, Xiao

doi:10.1177/1550147720971504

Cited by 19 publications

(6 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Given the labor-intensive process of labeling trajectories, a growing number of researchers have turned their attention towards exploring unsupervised learning methods [33]- [35]. These methods aim to learn patterns and detect outlier trajectories without the demands for pre-labeled data.…”

Section: B Learning-based Methodsmentioning

confidence: 99%

Transfer Learning for Region-Wide Trajectory Outlier Detection

2023

IEEE Access

View full text Add to dashboard Cite

Trajectory outlier detection is a crucial task in trajectory data mining and has received significant attention. However, the distribution of trajectories is tied to social activities, resulting in extreme unevenness among regions. While existing methods have demonstrated excellent performance in regions with sufficient historical trajectories, they frequently struggle to detect outliers in regions with limited trajectories. Unfortunately, this issue has not received much attention, leaving a gap in the current understanding of trajectory mining. To deal with this problem, we in this paper propose a model called TTOD that can effectively detect outliers in regions with sparse data by transferring knowledge among regions. The main idea is to learn a feature mapping function that maps the global feature space of auxiliary regions to the target region's specific feature space. To achieve this, we adopt a VAE-based model called the Global VAE to learn the global feature space in auxiliary regions by modeling the trajectory patterns with Gaussian distributions. Then, we propose a Specific-region VAE that serves as the mapping function to learn the target feature space. Additionally, considering the data drift of feature distributions among regions, we introduced an additional pattern synthesis layer, named the De-drift Layer, to diversify the target feature space, thus addressing the pattern missing issue caused by the gap of feature distributions between the auxiliary regions and the target regions. Then the target feature space can be well studied and applied to detect outliers. Finally, we conduct extensive experiments on two real taxi trajectory datasets and the results show that TTOD achieves state-ofthe-art performance compared with the baselines.INDEX TERMS trajectory outlier detection, transfer learning, VAE, spatial-temporal data, trajectory data mining.

show abstract

Section: B Learning-based Methodsmentioning

confidence: 99%

Transfer Learning for Region-Wide Trajectory Outlier Detection

2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In [18] investigated the use of an interactive IoT cloud deep learning model to develop a personalized learning system for university students. The authors found that the model was effective in identifying individual learning needs and could be used to develop personalized learning resources that improved learning outcomes.…”

Section: Related Workmentioning

confidence: 99%

Interactive IoT Cloud Deep Learning Model for Research Development in Universities for the Educational Think Tank

Jiang,

Liang

2023

IJRITCC

View full text Add to dashboard Cite

The construction of university education think tanks using the interactive service platform enables the sharing of research resources, encourages cross-disciplinary research collaboration, and fosters innovation in education. It also helps to build a stronger relationship between academia and industry by enabling practitioners to participate in research activities. The Internet of Things (IoT) can be used to collect and analyze data from various sources, including sensors and other connected devices, to provide insights into education-related issues. The integration of these technologies in university education thinks tanks can help to enhance the efficiency and effectiveness of research, decision-making, and collaboration processes. Hence, this paper constructed an Interactive IoT Cloud Computing Platform (IIoTCC). With IIoTCC model the innovative idea about research and other ideas are collected and stored in a Cloud environment. Within the environment, information collected is stored in the stacked architecture model with the voting-based model. Through the stacked model, information is processed and evaluated for academic activities. The IoT environment is implemented through IIoTCC for the information process in a deep learning environment for academic-related issues. Simulation analysis expressed that IIoTCC model achieves a higher accuracy rate of 99.34% which is significantly higher than conventional classifiers.

show abstract

“…In this context, FDIA detection is treated as a supervised binary classification problem. Based on the research work done in [14] the SVM is more efficient than CNN and KNN in anomaly detection with 91.29% of precision. Moreover, the SVM is a popular practice for training a decision boundary that divides data into several classes.as shown in Fig.…”

Section: Proposed Modelmentioning

confidence: 99%

Cyber-Attack on P2P Energy Transaction Between Connected Electric Vehicles: A False Data Injection Detection Based Machine Learning Model

2022

View full text Add to dashboard Cite

When cybersecurity is neglected, any network system loses its efficiency, reliability, and resilience. With the huge integration of the Information, Communication and Technology capabilities, the Connected Electric Vehicle (CEV) as a transportation form in cities is becoming more and more efficient and able to reply to citizen and environmental expectations which improve the quality of citizens' life. However, this CEV technological improvement increases the CEV vulnerabilities to cyber-attacks resulting to serious risks for citizens. Thus, they can intensify their negative impact on societies and cause unexpected physical damage and economic losses. This paper targets the cybersecurity issues for CEVs in parking lots where a peer-to-peer(P2P) energy transaction system based on blockchain, and smart contract scheme is launched. A False Data Injection Attack (FDIA) on the electricity price and power signal is proposed and a Machine Learning/SVM classification protocol is used to detect and extract the right values. Simulation results are conducted to prove the effectiveness of this proposed model.

show abstract

Unsupervised learning trajectory anomaly detection algorithm based on deep representation

Cited by 19 publications

References 29 publications

Transfer Learning for Region-Wide Trajectory Outlier Detection

Transfer Learning for Region-Wide Trajectory Outlier Detection

Interactive IoT Cloud Deep Learning Model for Research Development in Universities for the Educational Think Tank

Cyber-Attack on P2P Energy Transaction Between Connected Electric Vehicles: A False Data Injection Detection Based Machine Learning Model

Contact Info

Product

Resources

About