Improving elasticity in cloud with predictive algorithms

Abstract

Cloud computing is an innovation that is of expanding request nowadays. Here, resources are multiplexed from physical machines to virtual machines through virtualization technology. Cloud computing gives different sorts of administrations to clients. In Cloud Computing, the supplier progressively distributes the resources. Doing as such, the service provider ought to have some information about the future asset needs. They can be determined utilizing the load prediction calculations. A calculation named long short-term memory (LSTM) neural system is utilized to analyze the load, which is proficient as far as both expanding and diminishing need of resources. The predicted results of the LSTM model is helpful for optimizing the service response time and also fulfils the Service Level Agreement (SLA) contracted by the user.