Overview: Decision makers are often confronted with complex problems arising from multiple sources (i.e. product, project, organization, and process). These problems are challenging to analyze due to the various dependencies and interactions between their parts (i.e., subsystems) or with the environment. Thus, there is a need to build intelligent tools to understand, analyze and solve these problems encountered in daily practice. One of the prevailing tools extensively used to analyze complex problems is the simulation model(s) (e.g., agent-based, discrete event and systems dynamics). Simulation models enable decision-makers to set-up “what-if” questions about the effects of alternative pre-specified decisions on the system’s performance. Nevertheless, simulation models do not suggest or find optimal decision(s) regarding the solution of these complex problems. Thanks to the recent increases in the computational power of computers, this shortcoming of simulation models has been alleviated by coupling them with computational intelligence (CI) techniques (e.g., evolutionary computing, artificial neural networks, fuzzy sets). This coupling enables decision makers to reveal the patterns and trends and find (or near) optimal decisions for complex problems. There have been growing use of simulation and CI coupling across a variety of domains including defense applications, environmental systems, communication networks, supply chains, and healthcare systems.
Scope: This special session seeks challenging applied, and methodological papers focused on any aspect of the interplay between computational intelligence and simulation modelling for complex decision-making problems. In particular, the case studies on integration of simulation models with CI are encouraged to submit. We also welcome articles in, but not limited to, the following areas:
Surveys on the coupling of simulation models and CI
Applications of simulation and CI in the area of decision support systems and decision making under uncertainty
Applications of state-of-art CI techniques (e.g., evolutionary computing) in simulation models
Embedding machine learning techniques (e.g., reinforcement learning) into simulation models