We present some of the transdisciplinary projects we are involved in.

Uncertainty quantification and robust decisions

Computational models play an ever-increasing role in informing decisions. Epidemiologists guide public mitigation efforts in the current pandemic by predicting the effect of social distancing rules on the disease’s spread. Economists advise central banks about their monetary policy by forecasting the impact of changes in the interest rate on economic activity. Financial institutions manage their capital requirements by conducting stress tests about their business viability under adverse market conditions. Domain-expertise is essential for developing models tailored to their intended application and the available data. However, the shared need to calibrate models to data and enable model-informed decisions creates many transdisciplinary research opportunities.

Uncertainty, for example, is a major challenge across scientific domains. Epidemiologists face substantial uncertainties about the transmission of disease. Economists struggle to pin down the transmission channels of monetary policy to real economic outcomes, and financial institutions make assumptions about rare events such as financial crises and bank runs. Needless to say, all computational models are subject to numerous sources of uncertainty. The model is always misspecified, there are numerical approximation errors in its implementation, and the parameterization remains partly uncertain. Uncertainty quantification is a systematic attempt to characterize, manage, and reduce uncertainty and it is long recognized as a cornerstone of sound computational modelling. An explicit treatment of the uncertainties is crucial to properly frame model-informed decision-making as a decision problem under uncertainty.

This project establishes an analysis pipeline for exemplary computational models from economics, epidemiology, and finance. We collaborate on developing new tools for proper uncertainty quantification and robust decision-making by tackling its computational challenges using robust optimization techniques and combining them with recent advances in copula and surrogate modelling.

Partners:  Philipp Eisenhauer, Jan Hasenauer, Lena Janys, Daniel Oeltz, Dilan Pathirana
Funding: Excellence Strategy—TRA Modelling

Transdisciplinary Research Portfolio

Transdisciplinary research requires a basic understanding of the challenges and current practices within each of the disciplines involved. This common ground reveals opportunities for intellectual arbitrage across disciplines and allows identifying shared research needs. Accessible use-cases and example projects are needed to establish a shared understanding of the relevant terminology and to remove barriers to collaborations. We set up an online presence that includes a website documenting our ongoing work in economics, epidemiology, and finance. We contribute several of our group’s research codes for the domain-specific computational models and software supporting their analysis. We showcase application examples from all our domains to illustrate their common structure. All included codes emphasize the use of sound software engineering practices to allow for their collaborative development. Mathematics is the shared language across domains. As such, the detailed documentation of the underlying mathematical models is an essential part of our effort.

Partners:  Philipp Eisenhauer, Jan Hasenauer, Lena Janys, Daniel Oeltz, Dilan Pathirana
Funding: Excellence Strategy—TRA Modelling

E2: Economics meets Epidemiology

Public policy interventions are critical for the management of pandemics, but predictions about their impact are often flawed as they disregard the behavioral responses by individuals. We will establish an integrative framework combining epidemiological and economic modelling to address this challenge. Using the insights and data generated by the current COVID-19 pandemic, we seek to inform decision making processes in current and future health challenges explicitly accounting for the pervasive uncertainties involved in the analysis.

Partners:  Philipp Eisenhauer, Jan Hasenauer, Lena Janys
Funding: Volkswagenstiftung