Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Interpretable intrusion detection for IoT: a CNN-BiLSTM permutation importance framework for deep feature selection.

Ibrahim Al-Shibly1, Llorenç Burgas1, Joaquim Massana1

  • 1Control Engineering and Intelligent Systems, Universitat de Girona, Campus Montilivi, EPS IV, Girona, Catalonia, Spain.

Frontiers in Big Data
|June 8, 2026
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Optimal scheduler for energy consumption reduction in multi-vector energy management systems: A case study at the Port of Borg.

Heliyon·2024
Same author

Signaleeg : A practical tool for EEG signal data mining.

Neuroinformatics·2021
Same author

Insulin Recommender Systems for T1DM: A Review.

Advances in experimental medicine and biology·2020
See all related articles

A novel deep feature selection (DFS) framework using CNN-BiLSTM enhances intrusion detection in Industrial Internet of Things (IIoT) networks. This method improves detection accuracy, especially for imbalanced datasets, by selecting relevant features for lightweight models.

Area of Science:

  • Cybersecurity
  • Machine Learning
  • Industrial Internet of Things (IIoT)

Background:

  • Industrial Intrusion Detection Systems (IDS) face challenges with complex network traffic and evolving attacks.
  • Traditional feature selection methods struggle with non-linear dependencies and class imbalance, impacting detection performance.

Purpose of the Study:

  • To propose a Deep Feature Selection (DFS) framework for enhanced IDS in IIoT environments.
  • To address limitations of traditional feature selection methods in handling complex, temporally correlated network data.
  • To improve detection accuracy and deployability in resource-constrained IIoT settings.

Main Methods:

  • Utilized a hybrid Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) model for DFS.
Keywords:
CIC IIoT 2025CNN–BiLSTMIIoT intrusion detectioncybersecuritydeep feature selectionmachine learning

Related Experiment Videos

  • Assessed feature importance using permutation importance with time-aware perturbations.
  • Trained lightweight traditional machine learning models on selected features.

    • Main Results:

    • The CNN-BiLSTM DFS framework demonstrated improved recall and F1-score compared to Mutual Information (MI).
    • Enhanced performance was particularly evident in class-imbalanced scenarios.
    • The approach achieved a balance between detection accuracy, robustness, and deployability.

    Conclusions:

    • The proposed CNN-BiLSTM DFS framework offers a robust solution for IDS in IIoT.
    • Decoupling feature selection from inference enhances real-world applicability in constrained environments.
    • This method effectively handles multi-feature, temporally correlated traffic and dynamic attack patterns.