Prognostic Analytics Framework for Streamlining Cloud Selection in Machine Learning Deployments

In the largely expanding landscape of cloud service providers, the selection of an optimal cloud platform has emerged as a critical consideration. It would offer potential cost saving and operational efficiency enhancements. In this study, a machine-learning framework for prognostic analytics is proposed. The main goal is to create a streamlined and effective procedure for choosing the best cloud service provider for machine learning model deployments. This approach seeks to greatly improve cloud selection decision-making and operational effectiveness by leveraging the power of machine learning model which is trained on large datasets of cloud service providers' cost. The model also factors in aspects like model run time, GPU usage, CPU usage, networking metrics and instance family, memory usage, reservation and on demand cost to choose an optimal service provider at a minimal cost.