Attack Detection Using an Ensemble Model with Improved Score-Level Fusion in Cloud Computing

An Intrusion Detection System (IDS) is a hardware or software application which is used to monitor policy violations or unnecessary activities involved in systems. Such activities are immediately reported either to a formed midway utilizing security data or a supervisor. There are various types of IDSs, but the two most common IDSs are Host Intrusion Detection System (HIDS) and Network Intrusion Detection System (NIDS). In this paper, the proposed technique named as Ensemble Classification-based Cyber-Attack Detection (ECCAD) model is involved with processes like pre-processing, feature extraction and attack detection. The SMOTE-ENC algorithm is used in an enhanced class imbalance process to perform the pre-processing stage first. In the second step, features like improved entropy-based features, mutual information, information gain and statistical features are explained and evaluated. Finally, the main process, attack detection, takes place by an ensemble classifier with an improved score-level fusion process. Therefore, the efficacy of the proposed ECCAD method is evaluated over other existing approaches regarding different performance measures. The ECCAD achieved a larger accuracy of 98.81%.