Enhancing intrusion detection systems using hybrid deep learning models
Parag Deoskar, Ajay Kumar Sachan · International Journal of Cloud Computing and Database Management · 2025
Intrusion Detection Systems (IDS) play a crucial role in safeguarding networks by detecting malicious activities and threats. This research leverages two popular datasets, KDD 1999 and UNSW-NB15, to build a robust IDS model. The process begins with data selection using the pandas package, followed by data preprocessing, which involves handling missing values, label encoding, and the application of the SMOTE technique for data balancing. Subsequently, the data is split into training and testing sets for model evaluation, with PCA (Principal Component Analysis) applied for dimensionality reduction. For classification, several deep learning algorithms are employed, including ResNet-50, ResNet-101, Long Short-Term Memory (LSTM), Bidirectional LSTM, and a hybrid model of ResNet-50, ResNet-101, and ResNet-34. The system aims to predict intrusion attacks, providing valuable insights into the security landscape. Performance is evaluated using accuracy, precision, recall, F1-score, and comparative graphical analysis to assess the model’s efficiency. This framework contributes to enhancing IDS by offering a scalable and accurate solution for intrusion detection.