The rapid advancement of machine learning and the intelligence it enables presents a unique opportunity to enhance systems analysis and its applications in policy, societal decision-making and practice implementation. ASA uses a variety of these techniques to convert data into information and knowledge, to make modelling of complex systems more efficient, and to enable new insights which cannot be produced by other techniques.
Selected Research Highlights:
- The NODES research group is harnessing the intersection of community science and machine learning to produce high-resolution data in areas such as poverty and pollution, in particular, where data is scarce. This research advances new approaches to merge machine learning applications with citizen science to fill data gaps.
- The SYRR research group has used machine learning methods to model risk and gauge the socio-economic consequences of extreme event impacts. It is exploring the use of AI techniques for generating advanced insight for measuring disaster and climate vulnerability as well as resilience.
- In the EM research group, we leverage machine learning, for example in participatory decision-making exercises designed to integrate models, stakeholders, and technology to collaboratively tackle complex, large-scale, multi-objective challenges. These exercises, known as Scenathons, utilize machine-learning techniques through the SmartLinker Platform, employing multi-agent reinforcement learning to coordinate independent country and regional models. We further explore integrating AI in agent-based models (ABMs) by replacing traditional rule-based decision-making with artificial intelligence. Here, we investigate how large language models (LLMs) can enable agents to reason contextually and replicate human-like economic behaviors more realistically.
Javalera Rincón, V. , Orduña-Cabrera, F. , Obersteiner, M. , & Rios, A. (2022). Smart Linker for harmonizing trade assumptions between countries' sustainable pathways on the FABLE's Scenathon. IIASA Report. IIASA, Laxenburg, Austria
Sieg, T., Schinko, T. , Vogel, K., Mechler, R. , Merz, B., & Kreibich, H. (2019). Integrated assessment of short-term direct and indirect economic flood impacts including uncertainty quantification. PLoS ONE 14 (4) e0212932. 10.1371/journal.pone.0212932.
Fraisl, D. , See, L. , Fritz, S. , Haklay, M., & McCallum, I. (2024). Leveraging the collaborative power of AI and citizen science for sustainable development. Nature Sustainability 10.1038/s41893-024-01489-2.
