both firms employ the identical price-setting algorithm based on Q-learning. Thus, the question arises whether the … underlying assumption that both firms employ a Q-learning algorithm can be supported as an equilibrium in a game where firms can … chose between different pricing rules. Our simulations show that when both firms use a learning algorithm, the outcome is …

uses a new contextual bandit algorithm, which we call MTGP-LR, to learn the reward functions that map features to trading … algorithm employs a new online change-point detection test to learn in non-stationary environments. As an application in optimal …