A Simulation of Bayesian Surprise

Despite extensive efforts in computational modeling of the emotion of surprise, the literature seems far from reaching a consensus. Our contribution is to provide in silico evidence for the theory of Bayesian surprise. We aim to further elucidate Bayesian surprise through its connection with the inductive theory of inference in a simple RPS experiment. It is then observed that the results explain the two main forms of surprise that happen when we are puzzled or when we are enlightened with unexpected data.