Belief networks as comples systems

Abstract

There has been extensive work on understanding belief, from a psycho-logical, philosophical and neurobiological perspective. Meanwhile, arti-ficial intelligence has produced compelling developments that can enrich and update the brain-as-a-computer metaphor and has tried to better represent beliefs as cognitive probabilistic processes. In parallel, there has been a surge of research in Complexity Sciences, with applications rang-ing from Medicine to Finance. Some authors have already linked the con-nected nature of belief to the behaviour of complex networks. We would like to expand this approach to understand belief as a complex system with the main functions of providing a model of the world – including the individual and her surroundings – and producing guidelines for action. The complex-system perspective allows us to understand some of the properties of belief systems in a comprehensive manner, which many authors have begun to study in isolation. Notably, this provides a framework to study the impor-tant phenomena of belief formation and change as processes of emergence and adaptation. In this exploratory paper, we propose an outline for this framework for this study.

Keywords

• artificial intelligence
• artificial neural networks (ann)
• belief
• belief adaptation
• belief networks
• complex systems
• decision making
• emergence
• pattern recognition
• reinforcement learning (rl)