MODÉLISATION NEUROCOMPUTATIONNELLE ET PRISE DE DÉCISION

redpill redpill

Introduction arrow

Decision making is constantly happening in one’s daily life. It is a process that consists in a great diversity of mechanisms, all of them based on a variety of neural substrates. In computational neuroscience, modelling these neural substrates allows a better understanding of the underlying processes. Decision making particularly seems to rely on knowledge and/or past experiences, which suggests that learning is strongly related to it. Therefore, it also appears interesting to model learning substrates in decision making; in particular Schultz & al. (1997) tried to model dopamine cells activity whose role is known, especially in reinforcement learning. Later, this dopamine cells activity’s model has been further studied by Pan & al. (2005).