A lightweight intrusion detection system for internet of vehicles based on transfer learning and MobileNetV2 with hyper-parameter optimization

Wang, Ying-Qing; Qin, Guihe; Zou, Mi; Liang, Yan; Wang, Guofeng; Wang, Kunpeng; Feng, Yong; Zhang, Zizhan

doi:10.1007/s11042-023-15771-6

Cited by 7 publications

(3 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Li et al [13] proposed an intrusion detection method for Internet of Vehicles based on transmission double-depth Q network. Wang et al [14] designed a lightweight intrusion detection method that employs MobileNetv2 as the backbone, integrating transfer learning techniques and hyper-parameter optimization methods.…”

Section: Intrusion Detection Methods Based On Traditional Machine Lea...mentioning

confidence: 99%

Intrusion Detection Model for Internet of Vehicles Using GRIPCA and OWELM

Zhang,

Yang,

Shao

et al. 2024

IEEE Access

View full text Add to dashboard Cite

With the rapid development of the Internet of Vehicles, a large amount of vehicle network data is being generated. The large amount of data presents network communication security challenges. Although intrusion detection technology can assist in safeguarding the system from malicious attacks, the substantial data generated within the vehicle network poses time-consuming detection challenges. Thus, we propose an intrusion detection model for the Internet of Vehicles, utilizing Gaussian random incremental principal component analysis (GRIPCA) and optimal weighted extreme learning machine (OWELM). First, we utilize GRIPCA to reduce data redundancy by projecting high-dimensional data into a low-dimensional space, thus reducing storage costs. Then, we utilize the dynamic inertia weight particle swarm optimization (DPSO) to optimize the parameters of the weighted extreme learning machine (WELM) to achieve the best performance. We utilize the NSL-KDD and CIC-IDS-2017 datasets to perform experiments and compare the results with other techniques. The experimental results show the excellence of the proposed model, achieving an accuracy rate of 91.02% on the NSL KDD dataset and 94.67% on the CIC-IDS-2017 dataset.INDEX TERMS Extreme learning machine, Internet of Vehicles, intrusion detection, particle swarm optimization, principal component analysis.

show abstract

Section: Intrusion Detection Methods Based On Traditional Machine Lea...mentioning

confidence: 99%

Intrusion Detection Model for Internet of Vehicles Using GRIPCA and OWELM

Zhang,

Yang,

Shao

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…5) Using the hyper-parameter tuning optimization approaches enables us to obtain the best possible model architecture for a given task or dataset, including the ideal number of layers, neurons, or connections that can achieve the optimal balance between model size, complexity, and accuracy. This approach can reduce the size and complexity of the model while simultaneously increasing its performance and efficiency [390].…”

Section: B Learning Model Design In Iiotmentioning

confidence: 99%

“…• Using the distillation technique to transfer the knowledge or functionality of a large or complex model (teacher) to a small or simple model (student) [389]. • Using hyperparameter tuning optimization approaches to obtain the best possible model architecture for a given task or dataset [390].…”

Section: F Model Drift In Iiot Networkmentioning

confidence: 99%

Handling Privacy-Sensitive Medical Data With Federated Learning: Challenges and Future Directions

Aouedi

Sacco

Piamrat

et al. 2023

IEEE J. Biomed. Health Inform.

View full text Add to dashboard Cite

Recent medical applications are largely dominated by the application of Machine Learning (ML) models to assist expert decisions, leading to disruptive innovations in radiology, pathology, genomics, and hence modern healthcare systems in general. Despite the profitable usage of AI-based algorithms, these data-driven methods are facing issues such as the scarcity and privacy of user data, as well as the difficulty of institutions exchanging medical information. With insufficient data, ML is prevented from reaching its full potential, which is only possible if the database consists of the full spectrum of possible anatomies, pathologies, and input data types. To solve these issues, Federated Learning (FL) appeared as a valuable approach in the medical field, allowing patient data to stay where it is generated. Since an FL setting allows many clients to collaboratively train a model while keeping training data decentralized, it can protect privacy-sensitive medical data. However, FL is still unable to deliver all its promises and meets the more stringent requirements (e.g., latency, security) of a healthcare system based on multiple Internet of Medical Things (IoMT). For example, although no data are shared among the participants by definition in FL systems, some security risks are still present and can be considered as vulnerabilities from multiple aspects. This paper sheds light upon the emerging deployment of FL, provides a broad overview of current approaches and existing challenges, and outlines several directions of future work that are relevant to solving existing problems in federated healthcare, with a particular focus on security and privacy issues.

show abstract