Open Access Open Access  Restricted Access Subscription Access

Invalid cloud providers' identification using support vector machine

Seyedeh Zeynab Mohammadi, Nima Jafari Navimipour

Abstract


Cloud computing is a relatively new technology by the creation of which, companies and organizations transmit their services to cloud domains in order to do their tasks. With the companies and service provider organizations, profiteers seem not to neglect these areas and with different approaches try to do threats on users’ information and data security. Hence, it is necessary to adopt an approach relying on cloud providers to distinguish valid from invalid cloud provider. Recognizing valid and invalid cloud providers can be issued as a classifying subject. In this paper, a support vector machine (SVM) algorithm is used in order to classify cloud providers. The features of SVM are good generalization, ability to classify input pattern, optimal general pattern, and learning capability. The proposed method converts data to learning vector each of which has a corresponding output value and the ability to find the optimal amount in the non-linear and linear atmosphere. In this research, a data set of 1018 samples was used to classify cloud providers each of which has 10 features of cloud providers. To evaluate the performance of the proposed approach, 80 percent of data set is randomly considered as a training set and 20 percentage as a test set. The results demonstrated that the proposed approach is efficient as compared to the meta-heuristic algorithms.

Full Text:

PDF