24. February 2021

On the way to better decision-making Better decision-making: researchers link models from economics, epidemiology and financial research

In a transdisciplinary project, researchers at the University of Bonn are linking computer-based models from economics, epidemiology and financial research

Making a decision involves risks and uncertainties - as everyone experiences in their everyday lives. In pandemic times, however, the complexity of decisions and the many uncertainties that come together from different areas become all too apparent. A team of economists, epidemiologists, mathematicians and financial experts has now joined forces at the University of Bonn to bring their individual models together and calculate uncertainties. The goal: Making decisions based on a wider range of information. The Transdisciplinary Research Unit "Mathematics, Modelling and Simulation of Complex Systems" at the University of Bonn is funding the project with nearly 100,000 euros over two years. The results may be relevant for policy advice.

The Graph
The Graph - represents the "price of complexity" of a model. The green dashed curve represents the model error, which is the sum of two opposing errors (orange and blue curve). The orange curve shows the error resulting from uncertainty. As models become more complex, more and more uncertainty is built up. © Graphics adopted by Andrea Saltelli/ https://www.nature.com/articles/s41467-019-11865-8
Download all images in original size The impression in connection with the service is free, while the image specified author is mentioned.

When epidemiologists use computer-based models to predict the effects of distancing rules on the spread of viruses, this comes with uncertainties such as the evolving incidence number. When economists calculate effects on markets in their models, for example, they have to deal with uncertainty about how price changes will become established. The forecasts of financial experts also have uncertainties. Currently, the coronavirus pandemic makes it clear: Major societal challenges and the complex questions they raise cannot be answered by any one scientific discipline alone. This is a thought that the University of Bonn took as an opportunity one and a half years ago to establish six university-wide so-called Transdisciplinary Research Areas with different thematic focuses in the course of the Excellence Strategy funding program.

"Dating platform" for researchers as the origin of the project

However, as in real life, it is not a given that researchers from different "worlds" but with similar interests will bump into each other. This is why the so-called Collaboration Platform - a kind of dating portal for researchers - has been established in the mathematically oriented Transdisciplinary Research Area "Modelling". It is where the project now being funded began.

"The platform helped us to overcome the boundaries of our own department and bring our project to maturity," says project coordinator Prof. Philipp Eisenhauer from the Institute for Applied Microeconomics at the University of Bonn. On board with him: Jun.-Prof. Lena Janys, a microeconometrician from the Institute of Finance and Statistics, who specializes in matters of health economics, and Prof. Jan Hasenauer from the Life & Medical Sciences Institute (LIMES). He is an expert in mathematical modeling of biological systems. Other colleagues gradually joined the team, including epidemiologist Dr. Dilan Pathirana from the LIMES Institute and mathematician and financial expert Dr. Daniel Oeltz from the Fraunhofer Institute SCAI in Sankt Augustin.

Computing and consolidating uncertainties

The commonality of the researchers involved: They all explore computational and statistical methods for modeling complex systems. Depending on the discipline, however, they rely on different models, each of which relates to a specific issue within their own discipline.

This means that uncertainties can only be insufficiently accommodated. The great common goal is to use a variety of methods to map and quantify the uncertainties in the individual models, in other words, to determine how relevant they are. "The resulting findings are robust, which means they can be used to make better-informed decisions," explains Lena Janys.

Over the next two years, the researchers want to compile as many models as possible into one large decision-theoretic model, which will be freely accessible in a kind of data archive. "Epidemiological models often do not incorporate behavioral adjustments, which in turn are a focus of economic modeling and computational finance models. So we benefit from each other," emphasizes Jan Hasenauer.

Additional 119,000 euros from the Volkswagen Foundation

The team's project has not only won them internal acclaim - since December of last year, the Volkswagen Foundation has been supporting the sub-project "E2: Economics meets Epidemiology" by Lena Janys, Philipp Eisenhauer and Jan Hasenauer as part of their funding program "Corona Crisis and Beyond" with 119,000 euros over 18 months.

More information about the project

The Transdisciplinary Research Area "Mathematics, Modeling and Simulation of Complex Systems”

The creation of models is at the heart of many quantitative sciences. Nowadays, computers are able to process increasingly large data sets in decreasingly short time to create scientific forecasting models. The understanding of how complex systems, with many components that interact with each other, actually work, is one of the biggest challenges of our time, arising in numerous areas of science and technology. In the Transdisciplinary Research Area (TRA) "Mathematics, Modeling and Simulation of Complex Systems" at the University of Bonn, researchers from a broad range of disciplines create models that not only describe complex systems, but are able to analyze their behavior. They do this using a combination of classical observational methods, data simulation and creative spirit.

Prof. Philipp Eisenhauer
Prof. Philipp Eisenhauer - from Institute for Applied Microeconomics at the University of Bonn © Simone Jost
Jun.-Prof. Lena Janys
Jun.-Prof. Lena Janys - from the Institute of Finance and Statistics at the University of Bonn © Barbara Frommann
Prof. Jan Hasenauer
Prof. Jan Hasenauer - from the LIMES Institute at the University of Bonn © Michael Haggenmueller
Dr. Dilan Pathirana
Dr. Dilan Pathirana - from the LIMES Institute at the University of Bonn © private
Dr. Daniel Oeltz
Dr. Daniel Oeltz - from the Fraunhofer Institute for Algorithms and Scientific Computing SCAI © Fraunhofer SCAI

Prof. Dr. Philipp Eisenhauer (Project coordinator)
Institute for Applied Microeconomics
University of Bonn
Phone: +49 228 73-9219
Email: peisenha@uni-bonn.de

Jun.-Prof. Dr. Lena Janys
Institute of Finance and Statistics
University of Bonn
Phone: +49 228 73-9183
Email: ljanys@uni-bonn.de

Prof. Dr. Jan Hasenauer
LIMES Institute
University of Bonn
Phone: +49 228 73-62336
Email: jan.hasenauer@uni-bonn.de

Dr. Dilan Pathirana
LIMES Institute
University of Bonn
Phone: +49 228 73-62264
Email: dilan.pathirana@uni-bonn.de

Dr. Daniel Oeltz
Business Area Computational Finance
Fraunhofer Institute for Algorithms and Scientific Computing SCAI
Phone: +49 2241 14-4224
Email: daniel.oeltz@scai.fraunhofer.de

Wird geladen