Abstract: Cloud computing is a distributed computing environment where huge number of computers are connected with a communication channel such as the Internet. It provides the facility to run an application on many linked computers at the same time.The rapid growth of power demand from business, and Web applications has led to the emergence of cloud-oriented data centers. Load prediction is a significant cost-optimal resource allocation and efficient energy saving approach for a cloud computing environment. Load classification before prediction improves prediction accuracy.With the increasing demand on cloud provisioning instantly developing Vm in the cloud is not possible; it takes several minutes which leads to latency. So, Load prediction can minimize the downtime of the Vm. various load prediction models and the approaches based on the models are listed out in the paper. Methods through which error rate can be found are also described so that the prediction techniques can be well compared. Also the areas that can be worked upon after the successful Load prediction are listed into the paper.

Keywords: Cloud computing, Load balancing Load prediction, prediction accuracy.