Epsilon Open Archive

Abstract

A central issue related to climate change and the path to a low carbon society is how we can change our attitudes and associated behavioral patterns. This type of decisions is concerned with how complex systems can be dealt with, conceptually, psychologically, as well as socially. In order to transform our society, we need to consider the relationship between brain, mind and behavior. One of the approaches to address this problem is to design computational models that can be used for simulations and scenario building.

This thesis concerns the development and application of a neurocomputational model of the decision making process of an individual at experiential and social levels, considering both emotional and rational aspects. It is an attempt to bridge the gaps between micro (neuronal), meso (brain areas) and macro (cognition/behavior) levels with a focus on the mesoscale neurodynamics of cortical structures. The model is intended to link neural structures, functions, and includes effects of internal and environmental factors.

The thesis is divided into two parts, corresponding to the two kinds of decision making: 1) experience-based and 2) social-based decision making. At an individual level, a final decision is the result of an integration of rational and emotional processes. The neural structures involved in cognition valuate the potential options regarding internal attitudes and rules, as well as external contexts. Decision values are based on neural properties of activity patterns associated with different actions. The option with the highest value is selected for in the decision making process. Human behavior is guided not only by subjective values and attitudes, but also by the perceived behavior of others. Learning from/about others through observation shapes our thoughts and behavioral patterns. The second part of the thesis deals with this social adaptive characteristic of an individual, where the dynamic changes of her behaviors are connected with trust. Traces of social influences on an individual’s decisions and social expectations (e.g. trust) have been observed in the rational and emotional brain structures and their functions.

While the neurocomputational model is based on anatomical and physiological data of the modeled brain structures, no real world data have been available for model validation. Yet, simulation results mimic EEG and fMRI readouts, which could be compared with experimental/clinical data, when available. Future work intends to provide such data, but currently the modeling can only provide insights in the neurodynamic interactions between brain areas involved in decision making.