Meta-Strategy for Multi-Time Negotiation: A Multi-Armed Bandit Approach

Multi-time negotiation, which repeats negotiations many times under the same conditions, is an important class of automated negotiation. We propose a meta-strategy that selects an agent's individual negotiation strategy for multi-time negotiation. We model the meta-strategy as a multi-armed bandit problem that regards an individual negotiation strategy as a slot machine and utility of the agent as a reward. Our meta-strategy takes an individual negotiation strategy according to the opponent's strategy, its own profile, and the opponent's profile. The experimental results demonstrate the effectiveness of our meta-strategy under various negotiation conditions.