Data mining has been used in various areas, for example crime agencies, retail industries, financial data analysis, telecommunication industry, biological, among government agencies, etc. Several application handle very delicate data. So these data remains secure and private. In data mining, Classification could be the one of the major task. Going back two full decades various privacy issues are occurs so that many conceptual and feasible solutions to the classification problem have been developed. Similarly daily cloud user is increment tremendously and they have a big possibility to process the offload the information an encrypted form. The information in the cloud has been in encrypted form, recent privacy preserving classification systems are not feasible. In this paper, our proposed hybrid method provides privacy -preserving classifier for encrypted data of relational database and also achieves the marginally better performance for extracting information using k-NN algorithm from encrypted data of relational databases. This paper describes AES encryption technique which is highly secure and efficient.