Towards effective allocation of resources in service-oriented systems

Abstract

The article focuses on the problem of computing resource management in systems offering data processing methods in service-based model. It proposes a new model for service requests processing based on a concept of dynamic service selection in such a way, which meets client’s requirements, while minimizing the service delivery cost for the service provider. The key innovation of the described approach is the utilization of machine learning methods for effective estimation of non-functional parameters and resource allocation decision-making.