Treffer: Malicious host detection and classification in cloud forensics with DNN and SFLO approaches.

The rate of using cloud service is increased in recent years. The service provided by cloud computing (CC) is pre-owned by various laptops, smartphones, desktop computers, and notebook users. Cloud service enable the authorization practice due to an increasing number of cloud service users. Cloud service employs different host to deliver service to the users. But some hosts may be malicious and steal the user's information or else it provides an unwanted file instead of original files to the user. In previous works, this malicious hosts are identified by site re-routing links, distinguishing file types and so on. The main impact of this malicious host is that it delivers infected data or files to the user or it divert the user to the non-requested data and files. In this paper, we focus on identification and classification of malicious hosts. The host list is examined to extract the features of malicious host by applying firefly algorithm. This identified features are then pre-processed by principal component analysis (PCA) method. The Deep Neural Network based Shuffled Frog Leap Optimization (DNN-SFLO) algorithm is a famous deep learning (DL) approach proposed to test the optimized weights of an identified features. DNN-SFLO accurately detects the malicious host, because the presence of malicious host may affect the cloud service. Performance of DNN-SFLO based host detection is compared with Naïve Bayes, Neural Network (NN), Artificial NN (ANN), Fuzzy C-Means (FCM), Fuzzy k-Nearest Neighbour (FKNN), Support vector machine (SVM). Implementation for this host detection process is carried out in python. The performance metrics taken to evaluate the effectiveness of DNN-SFLO is F-measure, precision, G-mean, sensitivity, error detection probability, and recall [ABSTRACT FROM AUTHOR]

Copyright of International Journal of Systems Assurance Engineering & Management is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)