An Efficient Intrusion Detection Scheme Using Revised Equality Constraints based Lagrange's Multiplier for Cloud Applications

Abstract

            Cloud computing has become a major part of the IT industry in recent years. The end-users via the Internet easily provide a framework that connects to powerful services and applications in the cloud. An intrusion detection system is the most common mechanism for detecting an attack and is inefficient to be deployed in the cloud for unknown attacks.It is challenging to observe and perceive malicious activity within any manipulative system or web, or cloud. In the existing method,the intrusion detection system is designed based on the Stacked Contractive Auto-Encoder (SCAE)classification algorithm. SCAE classificationalgorithmin existing work requires high-dimensional representations method whichprovidesa deprivedpresentation for massive datasets.In the proposed scheme,the dataset is preprocessed using the Improved Feature Scaling (IFS) algorithmwith effective rules. The resultant preprocessed data is then trained with a training process. The trained information is then compared with the test dataset, and then the classification processis done using Revised Equality Constraints Based Lagrange's Multiplier (RECLM). The classification using RECLMprovides an outstanding classification result of 99.68% accuracy for attack typedetection for large datasets and multiple groupings in the clouds.