ABSTRACT
Cloud computing offers cost-effective resources on demand, elasticity, flexibility, availability, disaster recovery, etc. The resource availability directly impacts the resource provisioning and workload assignment. A cloud data center is equipped with a large but finite number of resource. For better user experience and workload management, these resources must be managed effectively. The workload estimation is a great help in achieving the aim of resource management. In this chapter, we have discussed the evolutionary neural networks and their role in workload forecasting. The both frameworks are trained using variants of the differential evolution learning algorithm which is one of the most reliable numerical function optimization algorithm. It uses evolutionary operators including mutation, crossover and selection to explore the solution space for an optimal solution.
