Pages: 58-66

Abstract

Cloud computing is a general term for referring to anything that requires the provision of services ‎hosted on the Internet. With advance in cloud computing, data error prediction has become an ‎important factor in cloud computing, so that predicting cloud errors is the most important barrier to the ‎speed and development of cloud computing software. The purpose of the study was to predict content ‎error in cloud computing based on perceptron neural network and RBF. In this research, a data set of ‎‎10,000 records has been used, including 23 features that were generally divided into three categories, ‎properties related to public security and properties related to the content of the data and the ‎characteristics required for cloud storage. In this research, the KFOLD method was selected as the ‎allocation of training and testing the data. The method of data selection was random. The software used ‎in this research was MATLAB, 2015. The results showed that the performance of the RBF system was ‎better than the other method in the two predictive systems‎‎.