An Anomaly-Based Intrusion Detection System (AIDS) built with a Random Forest classifier on the CICIOT23 dataset. This project automates the full ML pipeline to detect anomalous IoT network traffic with 99.76% accuracy.
SecureIoT–AI evaluates how DNNs and CNNs generalize to unseen IoT cyber-attacks under severe class imbalance using GAN-based augmentation, SMOTE, and class weighting.
Privacy-Preserving Intrusion Detection System for IoT Networks using Hybrid Deep Learning, Federated Learning, and Homomorphic Encryption
Robust and Privacy-Preserving IoT Intrusion Detection. Extends DL-BiLSTM (Wang et al., 2023) with class-weighted loss, FGSM/PGD adversarial hardening, and FedAvg federated learning. Achieves 80.72% accuracy on CICIoT2023 (8-class) with 75% adversarial robustness retention under FGSM.
Zero-day attack detection for IoT with six one-class models trained on benign CICIoT2023 traffic only and deployed on a Khadas VIM4 edge gateway. Includes training notebooks, trained models, on-device measurements, and a script that regenerates every table in the paper.
CICIoT2023 veri seti kullanılarak geliştirilen, paket-görüntü görselleştirme tabanlı ve %94,90 doğruluk oranına sahip hibrit CNN-Random Forest IoT saldırı tespit modeli.
Machine learning-based IoT intrusion detection using Random Forest trained on the CICIoT2023 dataset.
Intrusion Detection in CICIoT2023 using Logistic Regression, Adaboost Classifier, Gradient Boosting And MLP Model
ML-powered IoT Intrusion Detection System operating over MQTT. Simulates a smart device fleet, injects real attack vectors and scores traffic in real time using a Dual ML Engine with a live cyberpunk dashboard.
Edge-vs-cloud evaluation of five ML intrusion-detection models using the CICIoT2023 dataset, including Raspberry Pi benchmarking.
A composite resilience framework for consumer IoT: quantifying exposure, detection, and recovery against adaptive adversaries.
Scalograms and spectrograms generated from the CICIoT2023 dataset for AI-based IoT traffic classification and reproducible cybersecurity research.
Federated IoT intrusion detection on CICIoT2023 under IID and non-IID client heterogeneity using FedAvg and FedProx.
Add a description, image, and links to the ciciot2023 topic page so that developers can more easily learn about it.
To associate your repository with the ciciot2023 topic, visit your repo's landing page and select "manage topics."