Modeling of placebo effect in stochastic reward tasks by reinforcement learning

Abstract

Several studies in the field of computational psychiatry focus on modeling diseases using reinforcement learning. This study aims to model through reinforcement learning using real data of stochastic reward with regards to the placebo effect. More specifically, we model the placebo effect using real data related to the non-invasive brain stimulation technique. In contrast to human data, the results of the machine learning approach showed no difference owing to the differences in the placebo effect.