CLOUD MALICIOUS THREAT DETECTION BY FEATURES FROM INTELLIGENT WATER DROP SET AND EBPN

Cloud increase strength of various organizations to work from any location and time. This flexibility lead to increase some security issue. As malicious programs take advantage of multiple entries in the cloud, so threat detection system need to be develop. This paper has developed a cloud malicious threat detection model which learns the behavior of different ideal and attack conditions. Requesting session on cloud has different feature set. So paper has filter those feature set by Intelligent Water Drop Set Algorithm (IWDA). Random feature set were developed in the algorithm for reducing the feature overload during training. Difference between the feature value act as soil in the algorithm for comparing two water drop sets. Filtered feature were passed in Error back propagation neural network where sigmoid function was used for training. Experiment was perform on UNSW-NB15 dataset and comparison result shows that proposed model has increase the threat detection accuracy by %.