Postdoctoral Research Fellow in High-Performance Social Simulation (TU Delft)

Postdoctoral Research Fellow in High-Performance Social Simulation (TU Delft)

Postdoctoral Research Fellow in High-Performance Social Simulation (TU Delft)

TU Delft has an opening for a postdoctoral research fellow at the intersection of social simulation and high-performance algorithms.  The goal of project is to develop methods to speed large-scale agent-based modelling (ABMs), for example by means of effective parallelization using High Performance Computing (HPC).  The use of machine learning for efficient ABM simulations will enable uncertainty and global sensitivity analysis of large ensembles of ABM runs.  The postdoc will use the HPC cluster newly established at TU Delft in 2021.

The successful candidate will work closely with faculty members at the TU Delft faculties of Technology, Policy and Management, and Electrical Engineering, Mathematics and Computer Science.  Working with Prof Tatiana Filatova, Dr Jan Kwakkel and Dr Neil Yorke-Smith, the candidate will study how theory-informed meta-modelling and learned surrogate models can accelerate agent-based models.

This two-year position as a part of the EU project TAILOR (Foundations of Trustworthy AI, Integrating Reasoning, Learning and Optimization, tailor-network.eu) in conjunction with the TU Delft Institute for Computational Science and Engineering (DCSE, tudelft.nl/cse).  The TAILOR network is funded by the European Union to build the capacity of the scientific foundations for trustworthy AI in Europe, by developing a network of research excellence centres leveraging and combining learning, optimization and reasoning.  The successful candidate will have access to the TAILOR network, including travel and mobility funds.

The position is at the intersection of social simulation and high-performance algorithms, and links to two ongoing EU projects (TAILOR and ERC SCALAR).

Deadline for applications: 31 December 2021. Apply here:

 https://www.academictransfer.com/en/306558/postdoc-high-performance-social-simulation/apply/#apply

 

Towards global-scale behavioural models of land use change: Online Symposium

Towards global-scale behavioural models of land use change: Online Symposium

Towards global-scale behavioural models of land use change: Online Symposium

Location: Virtual
Dates: Wednesday 15 & Thursday 16 December 2021, 15:00-17:00 Central European Time

Register: https://docs.google.com/forms/d/e/1FAIpQLSd8n_KTfIQUn6zRe0VIBi8P_loe3OeIR8iiKxAptGPhjKPLhw/viewform

 

Organisers: Dr Calum Brown, Prof. Tatiana Filatova, Dr Birgit Müller, Dr Derek Robinson, Prof. Mark Rounsevell, Prof. Maja Schlüter

This symposium will gather together an international group of researchers to tackle one of the most important challenges facing Earth System Science: the development of large-scale (national/continental to global) land-use models that are based on human behaviour, agency and decision-making processes. Participants will work together to support the development of such models, building on recent advances in theoretical and computational representations of human decision-making to determine which processes and relationships are most important, and how they can best be modelled. The workshop will introduce recent work within the behavioural modelling community, and develop specific collaborative efforts leading towards global-scale modelling. Participants are invited to present flash talks and ideas for group discussion, and to complete a brief pre-workshop survey. See the agenda below.

Large Scale Behavioural Modeling Symposium

Large Scale Behavioural Modeling Symposium

Large Scale Behavioural Modeling Symposium

 

 

Symposium: Towards global-scale behavioural models of land use change 15-17 December 2021, Schloss Herrenhausen, Hannover

Organisers: Dr Calum Brown, Prof. Tatiana Filatova, Dr Birgit Müller, Dr Derek Robinson, Prof. Mark Rounsevell, Prof. Maja Schlüter

Overview
The symposium will take place between 09:00 on Wednesday 15 December and 15:30 on Friday 17 December 2021. The symposium (as well as meals) will be hosted in Schloss Herrenhausen, and we will stay in a hotel in the centre of the city. Accommodation and meals will be provided, and there is limited travel funding for participants unable to cover their own expenses.

General description
This symposium will gather together an international group of researchers to tackle one of the most important challenges facing Earth System Science: the development of large-scale (national/continental to global) land-use models that are based on human behaviour, agency and decision-making processes. Participants will engage with theories, methods and tools that can support this development across the land use modelling community. The ultimate aim of the work initiated here will be to identify pathways to sustainability that account for fundamental processes in human and natural systems in uncertain future conditions. As such, these pathways are not detectable through current modelling approaches that do not represent the relevant processes. Participants will work together to support the development of a common land system ‘modelling framework’. This framework will build on recent advances in theoretical and computational representations of human decision-making to determine which processes and relationships are most important, and how they can best be modelled. In creating an alternative to the current range of ‘top-down’ global models based on macro-economics, the symposium will contribute to the development of a working laboratory to test theories of human decision making, assess the impacts of different forms of decision-making on land system outcomes, and undertake social simulation experiments in a globally connected world. This includes new representations of institutional processes and their relationships with local land users. We also envisage coupling of large-scale, land-use models with other model types, such as Dynamic Global Vegetation Models (DGVMs), biodiversity models and/or climate emulators. These coupled models can then be used to explore a wide range of environmental change drivers and to evaluate the consequences of these for ecosystem services. The symposium will involve a mix of research presentations, open discussions, workshops and informal time for networking and planning.

Purpose
The main purpose of the workshop will be to build on recent work within the behavioural modelling community to plan specific collaborative efforts leading towards global-scale modelling. To this end, participants will be invited to complete a pre-workshop survey on their experience and ideas for developing joint work, and to briefly present their own work in flash presentations. We will also try to learn lessons from other fields facing similar challenges. Survey responses will be developed during the workshop to produce a ‘ways forward’ paper and to inform development of the modelling framework.

Contributions
Flash talk and poster submissions are welcome. Flash talks will last 10 minutes with 10 minutes of questions. Talks and posters should be oriented around the symposium’s aims, and be relevant to large (continental-global) scales. If you would like to present a talk or poster, please send a title and short outline (100 words) by 29 October 2021.
We will also ask participants beforehand to prepare one page/slide on potential starting points for collaboration based on their current projects, and to answer a few key questions about research achievements and priorities.

Outputs
We hope to achieve a number of outputs from the symposium, including development of the
large-scale behavioural modelling network and specific plans for producing the common
modelling framework. Further outputs are anticipated to include a paper on ways forward for large-scale behavioural modelling and research funding proposals.

 

Read more about the symposium here.

Land Use and Ecosystem Change Summer School

Land Use and Ecosystem Change Summer School

Land Use and Ecosystem Change Summer School

An AIMES and GLP associated summer school on Land Use and ecosystem change ran between the 9th and 20th of August at the Karlsruhe Institute for Technology in Garmisch-Partenkirchen, Germany. Twenty-six Masters and PhD students from twelve different countries participated in the 2-week online course, covering a wide range of issues related to ecosystem functioning, socio-ecological systems and land use change.  The summer school involved a mix of webinars, practical exercises and student presentations. Having learned about biophysical and human processes, analysis and modelling techniques, the students went on to explore the role of research in international assessment processes such as the IPCC and IPBES. The summer school is an annual event and is planned to run as an in-person or hybrid event in August 2022. See landchange.earth/online-summer-school-2021 for more information.

Webinar materials now available for “What can land use modellers learn from other disciplines?”

Webinar materials now available for “What can land use modellers learn from other disciplines?”

The joint AIMES/GLP Working Group on Large-scale Behavioural Models of Land Use Change held its second webinar in April 2021 to discuss exciting new advances in social simulation and computational modelling that can support better representation of human behaviour in the land system. 

FirstRobert Axtell presented new work on large-scale agent based models in social science:

The U.S. private sector consisted, before the pandemic, of somewhat more than 120 million employees working in 6 million firms, making nearly $30 trillion of gross output, with another $1 trillion of output coming from firms that had no formal employees. Using data on all of these entities this research describes a model of the entire private sector, i.e., at 1-to-1 or full-scale (aka digital twin, mirror world) that closely reproduces dozens of gross statistical features of American firms and workers, things like firm sizes and ages, employee tenure and turnover, high rates of new firm entrance and endogenous firm exit, a variety of inter-firm networks, and so on. A book on this subject, Dynamics of Firms from the Botton Up: Data, Theories, and Models, is forthcoming from MIT Press.

Presentation: Agent-Based Modeling of Economic Phenomena at Very-Large (Full) Scale (PDF, 11 MB)

Papers:

SecondBrian Mac Namee described the use of machine learning and ABM in computer game design:

For almost every type of model we build we need to make decisions about the level of details with which we will model the real world in a virtual model. Badler (1997) refers to this as virtual fidelity. In the talk this was demonstrated in game agent models for enhancing player experience, machine learning models for behaviour prediction, and an agent-based model for modelling infectious disease transmission.

Badler, Norman I. “Real-time virtual humans.” Proceedings The Fifth Pacific Conference on Computer Graphics and Applications. IEEE, 1997.

Presentation: How Much Reality is Enough? Games, Cows and Disease (PDF, 7 MB)

Papers: 

  • O’Sullivan, Carol, Justine Cassell, Hannes Vilhjalmsson, J. Dingliana, Simon Dobbyn, Brian McNamee, Christopher Peters, and Thang Giang. “Levels of detail for crowds and groups.” In Computer Graphics Forum, vol. 21, no. 4, pp. 733-741. https://onlinelibrary.wiley.com/doi/full/10.1111/1467-8659.00631
  • Ryan, C., Gúeret, C., Berry, D., Corcoran, M., Keane, M. T., & Mac Namee, B. (2021). Predicting Illness for a Sustainable Dairy Agriculture: Predicting and Explaining the Onset of Mastitis in Dairy Cows. In Proceedings of Explainable Agency in Artificial Intelligence Workshop @ AAAI 2021 https://arxiv.org/abs/2101.02188
  • Hunter, Elizabeth, Brian Mac Namee, and John Kelleher. “A Hybrid Agent-Based and Equation Based Model for the Spread of Infectious Diseases.” Journal of Artificial Societies & Social Simulation 23.4 (2020). https://arrow.tudublin.ie/sciendoc/241/
  • Hunter, Elizabeth, Brian Mac Namee, and John Kelleher. “An open-data-driven agent-based model to simulate infectious disease outbreaks.” PloS one 13.12 (2018): e0208775. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0208775

ThirdThomas Clemen presented a simulation platform capable of modelling vast multi-agent systems:

The MARS Group at Hamburg University of Applied Sciences, Germany, provides a software framework for creating agent-based simulation models on a very large scale. The presentation introduces some principal aspects from a computer science and system theoretical perspective. Additionally, the challenges and opportunities of interdisciplinary teams were discussed. A short-course format to develop the necessary skills was proposed.

Presentation: Virtual Humans on MARS: Concepts, AI, and Interdisciplinarity (PDF, 2 MB)

Papers:

The webinar concluded with a discussion among presenters and attendees about how these advances can be used to develop a new generation of land system science models.

Become a Member
Join the Large-scale Behavioural Models of Land Use Change working group to be kept updated on the latest news, events and jobs of interest!

You can also view the recording and download materials from the first webinar.

Large scale behavioural models webinar: The role of machine learning, game design & parallelization in the future of land use modelling

Large scale behavioural models webinar: The role of machine learning, game design & parallelization in the future of land use modelling

Webinar: Friday, 30 April 2021 at 3:00 pm (CEST)
Register to receive Zoom connection information

The AIMES/GLP Working Group on Large-scale Behavioural Models of Land Use Change is pleased to invite you to their next webinar on 30 April at 15:00 Central European Summer Time.

In this webinar you will hear about exciting new advances in social simulation and computational modelling that can support better representation of human behaviour in the land system. Prof. Robert Axtell will present new work on large-scale agent based models in social science, Dr. Brian Mac Namee will describe the use of machine learning and ABM in computer game design, and Prof. Thomas Clemen will present a simulation platform capable of modelling vast multi-agent systems.

The webinar will conclude with a discussion among presenters and attendants about how these advances can be used to develop a new generation of land system science models.

Please note: Attendance is limited to the first 100 participants who join the meeting the day of the event, regardless of registration order. We will send the webinar materials and a recording link to all registrants after the event, regardless of attendance.

Please register here.

You can also view the recording and download materials from the first webinar.