Self-adaptive organizations for distributed search: The case of reinforcement learning

Abstract

In this paper we study the effects of learning by reinforcement and adaptive change of distributed search systems’ organizations. We find that employing learning by reinforcement to direct organizational alterations of distributed search systems may lead to high levels of systems’ performance and this, in particular, with rather high efficiency in terms of effort of reorganization. The results also suggest that the complexity of the search problem together with the aspiration level, relevant for the positive or negative reinforcement, considerably shape the effects of learning.