Secure and Efficient Regression Analysis Outsourcing Algorithm for Cloud Server

IT industry has been experiencing the benefits of outsourcing for years, since the outsourcing brings down both the capital and operational expenditure. Similarly, cloud computing provides storage, computation, and other specialized services on demand to customers over the internet at a very generous cost. However, outsourcing data and computation to a third party cloud server is a great cause of concern to the client because a customer physically loses control over their sensitive/classified data and computation. The loss of physical control over the data and computation is the main issue for a client that makes him feel insecure using cloud computing services. The solution to address outsourcing issues, first, the cloud service provider must be honest by providing correct and secure computation; second, the outsourced data and computation shall be verifiable to customers in terms of confidentiality and integrity. In this paper, we investigate the problem of regression analysis, outsourcing problem and devise a secure and efficient outsourcing algorithm which provides security for the input and also provide safeguards to the output result computed on cloud servers. Further, a novel and efficient result verification technique have been developed to detect server misbehavior and cheating with optimal probability of 1. Furthermore, theoretical and experimental analysis has compared with existing algorithm to demonstrate the efficiency, security and effectiveness of the proposed algorithms.