Delegating via Quitting Games

TL;DR

This paper introduces delegation policies based on quitting games and multi-armed bandits, effectively guiding recursive delegation decisions and outperforming non-recursive approaches.

Contribution

It presents novel delegation policies utilizing quitting games and bandit algorithms to handle recursive task delegation scenarios.

Findings

Quitting game-based policies outperform non-recursive methods.

Policies effectively manage recursive delegation interactions.

Empirical tests show improved delegation success rates.

Abstract

Delegation allows an agent to request that another agent completes a task. In many situations the task may be delegated onwards, and this process can repeat until it is eventually, successfully or unsuccessfully, performed. We consider policies to guide an agent in choosing who to delegate to when such recursive interactions are possible. These policies, based on quitting games and multi-armed bandits, were empirically tested for effectiveness. Our results indicate that the quitting game based policies outperform those which do not explicitly account for the recursive nature of delegation.