Extended Sparse Transient Search Deep Transfer Learning Based Intrusion Detection System

Abstract

Intrusion detection system (IDS) is typically responsible for tracking and identifying fraudulent behaviours in any operating network. This paper propose an extended sparse transient search deep transfer learning-based intrusion detection system (STSDTL-IDS) that overcomes the limitations and classifies complex attacks accurately and finely. A feature selection method, namely, chimp optimization algorithm (COA) is used to eliminate the unwanted features and it is used to detect assaults by identifying relevant aspects in the dataset with high precision. The proposed method is a hybrid method which includes sparse transient auto encoder with deep transfer learning, and extended Transient Search Optimization algorithm to improve the cloud IDS efficiency. The performance of the sparse deep transfer learning algorithm is improved using the extended transient search optimization. The python tool is used and the UNSW-NB15 and CICIDS2017 datasets are used. The experimental result shows that the proposed method outperforms when compared to the existing methods.