×C1. Complex systems modelling and artificial intelligence

Complex adaptive systems are collections of interacting agents that exhibit non-trivial collective behaviour. They have many applications in studying major societal and environmental challenges with diverse multi-scale interacting processes. In particular, complex systems analysis mainly focuses on modelling the micro-level processes that govern emergent behaviour, quantitively. As the intrinsic complexity of the problem grows, it becomes more difficult to extract the hidden micro-dynamics by process-driven modelling that relies on theories, experience, and human intuition with various prior assumptions. Addressing these limitation, artificial intelligence (AI) techniques have recently shown phenomenal performance in accurate predictions, pattern recognition, and function approximation. For example, reinforcement learning algorithm enables agent-based models to generate intelligent agents that learn how to optimize their behaviour according to the changes in the environment and reward structure dynamically.

Having said this, AI algorithms cannot estimate the emergent behaviours and do not include an adaptive component allowing micro behaviours to evolve in response to some environmental stimuli, which are necessary for creating a diversity of behaviours. For instance, the performance of AI in different circumstances can be only assessed when there is a simulation model to replicate the potential future scenarios. Hence, complex adaptive systems and AI present complementary capabilities that provide many promising opportunities to achieve the best insight into systems. This integrative approach is however still under development and opens several directions for future work.

In this session, we provide a forum for researchers to highlight and present their latest developments in the field of complex systems and artificial intelligence technologies, frameworks, architectures, algorithms, and applications. Research fields could include epidemiology, energy, traffic, biology, agriculture, economy, environment, management, etc. Both theoretical and experimental contributions containing novel applications with new insights and findings in the field are welcome.

Key topics: Machine learning, Integrative modelling, Rule inference, Complex adaptive systems