IoTSecUT: Uncertainty-Based Hybrid Deep Learning Approach for Superior IoT Security Amidst Evolving Cyber Threats
Bibliographic record
Abstract
The rapid expansion of digital infrastructure has led to increased security threats. Deep Learning (DL) algorithms have emerged as potent tools for detecting cyberattacks in IoT ecosystems. However, challenges like imbalanced class distribution and vast data volumes remain, resulting in inaccurate classification outcomes and inflated accuracy rates. Additionally, memory-constrained IoT devices often find it challenging to accommodate robust DL methods. This paper introduces an innovative approach that addresses class imbalance and the challenges posed by high-dimensional data in intrusion detection systems. Our framework encompasses: (1) a conditional Generative Adversarial Network (cGAN) for minority class upsampling, (2) an auxiliary autoencoder for dimensionality reduction, and (3) an unique hybrid uncertainty-based transformer architecture for efficient network traffic classification. We undertake extensive experiments on two specific datasets: BoT-IoT and CICIDS2018, affirming the efficacy of our hybrid DL-based approach. Initially, we assess the quality of synthetic data produced by various techniques, comparing their Principal Component Analysis (PCA) plots to authentic data. Our GAN-generated data closely mirror the PCA of real data, denoting a high similarity in distribution. Subsequently, we benchmark our auxiliary autoencoder against established dimensionality reduction techniques. The results indicate that our auxiliary autoencoder has significantly lower noise levels than contemporary methods, reducing the data storage volume of network traffic by 93.02% on BoT-IoT and 96.25% on CICIDS2018 datasets. Lastly, we illustrate the enhanced capability of our uncertainty-based attention model in detecting cyberattacks across both datasets.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".