Why Imitate , and if so , How ? A Bounded Rational Approach to Multi - Armed
نویسنده
چکیده
We consider the situation in which individuals in a ̄nite population must repeatedly choose an action yielding an uncertain payo®. Between choices, each individual may observe the performance of one other individual. We search for rules of behavior with limited memory that increase expected payo®s for any underlying payo® distribution. It is shown that the rule that outperforms all other rules with this property is the one that speci ̄es imitation of the action of an individual that performed better with a probability proportional to how much better she performed. When each individual uses this best rule, the aggregate population behavior can be approximated by the replicator dynamic. JEL Classi ̄cation Number : C72, C79
منابع مشابه
Variable Structure Behavioural Controller for Multi-agent Systems
In previous papers authors have considered agents as inertia-less self driven particles and designed a flocking algorithm. Application of this algorithm to agents with considerable inertial characteristics needs a behavioural controller. The controller uses the local information and helps every agent to imitate the desired behaviour as a member of the flocking frame which covers the main is...
متن کاملThe Effects of Mystical Monotheism on the Concept and Approach of Moral Practices
One of the important questions in the field of moral attributes and behavior is what changes the monotheistic view makes in human morality and behavior? Can it change the meaning of actions and deepen them? Can it modify and improve their approach? If so, how and why do such changes occur? This study tries to explain the effects of mystical monotheism in the concept and approach of moral practi...
متن کاملBeyond the Hazard Rate: More Perturbation Algorithms for Adversarial Multi-armed Bandits
Recent work on follow the perturbed leader (FTPL) algorithms for the adversarial multi-armed bandit problem has highlighted the role of the hazard rate of the distribution generating the perturbations. Assuming that the hazard rate is bounded, it is possible to provide regret analyses for a variety of FTPL algorithms for the multi-armed bandit problem. This paper pushes the inquiry into regret ...
متن کاملBounded regret in stochastic multi-armed bandits
We study the stochastic multi-armed bandit problem when one knows the value μ(⋆) of an optimal arm, as a well as a positive lower bound on the smallest positive gap ∆. We propose a new randomized policy that attains a regret uniformly bounded over time in this setting. We also prove several lower bounds, which show in particular that bounded regret is not possible if one only knows ∆, and bound...
متن کاملWhy and How Is Compassion Necessary to Provide Good Healthcare? Comments From an Academic Physician; Comment on “Why and How Is Compassion Necessary to Provide Good Quality Healthcare?”
This is a short commentary to the editorial issued by Marianna Fotaki, entitled: “Why and how is compassion necessary to provide good quality healthcare.” It introduces the necessity of a more cognitive approach to explore further the determinants of behavior towards compassionate care. It raises questions about the importance of training towards a more patient-care and values driven healthcare...
متن کامل