Prescriptive Analytics: When Data- and Simulation-based Models Interact in a Cooperative Way

Abstract

Business analytics is an extensive use of data acquired from diverse sources, statistical and quantitative analysis, explainable and predictive models, and fact-based management to make better strategic decisions for different stakeholders. To be able to model complex systems holistically in such a way that they can be fed into an efficient simulation-based optimization in the sense of prescriptive analytics, approaches and solutions that go beyond state-of-the-art are required. This paper introduces the basic technologies used in prescriptive analytics and proposes secure prescriptive analytics (SPA) that is based on component-based hierarchical modeling and dynamic optimization. Each element under the SPA framework is defined and illustrated by an example of production plan optimization.