Prescriptive Analytics:  When data- and simulation-based models interact in a cooperative way

Michael Affenzeller

University of Applied Science Upper Austria

Abstract:

Prescriptive Analytics is an interdisciplinary topic in an interdisciplinary field, or put another way it is a synergistic hybridisation of various methods and algorithms from statistics, computer science, artificial intelligence, mathematics and operations research. Its aim is to provide optimized recommendations for action in various application areas. In this way, knowledge gained in the digital world is brought back to the real world, providing better and more efficient procedures, designs and processes.

Simulation and data-based models have several advantages and limitations. Simulation-based representations of reality can be very accurate; however, they are often too costly for the purpose of integration into powerful optimization methods. Data-based surrogate models on the other hand can be highly efficient in application – but may not be as accurate. Appropriate hybridizations of these approaches in terms of prescriptive analytics have the potential to support and complement each other and decisively advance research areas such as simulation-based optimization.

The presentation will include methodological research topics that are currently being pursued in the HEAL research group [https://heal.heuristiclab.com/] led by Affenzeller as well as concrete project results that have already found their way into economic and industrial applications.

Short Bio:

Michael Affenzeller is professor for heuristic optimization and machine learning at the University of Applied Science Upper Austria and head of the master degree program for Software Engineering. He graduated in mathematical computer science and did his doctorate and habilitation in computer science with a special focus on applied systems science. Michael Affenzellers dissertation and habilitation were about algorithm development in the field of metaheuristic algorithms. He is heading the research group HEAL (Heuristic and Evolutionary Algorithms Laboratory) with around 20 FTE members. HEAL has been successful in hosting three Josef Ressel centers, among them Austria’s very first Ressel center Heureka!.