IJSR - International Journal of Scientific Research

K.PALANISAMY; DR.C.CHANDRASEKAR

Volume : VI, Issue : XI, November - 2017

ENSEMBLE OF RANDOM DECISION TREES WITH BOOTSTRAP AGGREGATING FOR IMPROVING CONFIDENTIALITY OF CLOUD DATA SERVICES

K. Palanisamy, Dr. C. Chandrasekar

Abstract :

Cloud computing is computing resources to deliver services using local servers or private devices in order to maintain the cloud applications. The cloud service provider is responsible for maintaining the cloud data on cloud database against the unauthorized access. Thus, cloud data security is one of main issue in a cloud computing environment with higher response time of cloud service provisioning. In order to overcome the above issues, Ensemble of Random Decision Trees with Bootstrap Aggregating (ERDT–BA) Technique is proposed. Thus, the proposed technique enhancing the security and confidentiality rate of cloud data service provisioning. In addition, Bagging Ensemble Classifier is used to improve the authentication performance of cloud users with minimum time. In Bagging Ensemble Classifier, Random Decision Trees is combined with Bootstrap Aggregating to attain higher classification accuracy for user authentication. With the aid Ensemble Classifier, cloud data is classified as legitimate or illegitimate users and verified cloud users only access the data. After cloud user verification, required data services are provided only to legitimate cloud users in cloud computing environment. This helps to attain secure cloud data service provisioning with higher integrity and confidentiality rate of cloud data services. The performance analysis of proposed ERDT–BA technique is conducted on parameters such as classification accuracy, false positive rate, and data confidentiality rate. The experimental result shows that the ERDT–BA technique achieves higher security and data confidentiality rate of cloud service provisioning than other methods.