Hybrid Machine Learning and Deep Learning-Based Intrusion Detection System for IoT Network Security Using Python

Abstract

The rapid expansion of Internet of Things (IoT) networks has increased the attack surface of connected systems used in smart homes, healthcare, transportation, industrial automation, agriculture, and critical infrastructure. IoT devices often operate with limited computation, memory, battery capacity, and security support, making them vulnerable to denial-of-service, botnet, probing, brute-force, spoofing, malware, and data injection attacks. Traditional signature-based intrusion detection systems are not sufficient for detecting unknown and evolving threats because they mainly rely on fixed attack patterns. This paper proposes a hybrid machine learning and deep learning-based intrusion detection system for IoT network security using Python. The proposed framework integrates data preprocessing, feature selection, machine learning classifiers, deep learning models, and hybrid decision fusion to improve attack detection accuracy and reduce false alarms. Random Forest, Support Vector Machine, Decision Tree, and XGBoost are considered as machine learning classifiers, while Deep Neural Network, Convolutional Neural Network, Long Short-Term Memory, and CNN-LSTM models are used for deep learning-based traffic analysis. The Python implementation uses Pandas, NumPy, Scikit-learn, TensorFlow/Keras, and Matplotlib for dataset handling, model training, performance evaluation, and visualization. The proposed hybrid ML-DL model is evaluated using accuracy, precision, recall, F1-score, false alarm rate, and confusion matrix analysis. The results demonstrate that hybrid fusion of classical machine learning and deep learning improves detection performance compared with individual models and provides a practical approach for real-time IoT intrusion detection.