Malicious Host Identification Using Machine Learning Classification In Cloud

G Nandita,G Kanika, T Munesh Chandra

Cloud services are used by countless users with different devices such as smartphones, notebooks, laptops and desktop computers. These uncountable users will get authorization to enter into the cloud network to access the cloud services. Cloud services are provided by the different host. Some of the hosts in the cloud network may provide Trojan, malware, and infected files or try to hijack user’s information. On related previous work, they identify a malicious host by differentiating file types, site redirection links and so on. The malicious host always sends an infected file to the users or it may try to redirect the user to access the non-requested files and data. In this project, we proposed a framework based on firefly algorithm to extract features from malicious host from the host list and principal component analysis (PCA) used for pre-processing. The cleaned data will be tested by using the popular classification technique convolutional neural network (CNN). And finally, the results will be classified by machine learning algorithm. Our work helps cloud forensics to identify the malicious host easily. MATLAB platform is used for simulation and the performance is estimated in terms of accuracy, standard deviation, and related work.

Volume 11 | Issue 7

Pages: 707-717