Learning Drifting Negotiations

Enembreck, Fabrício; Avila, Braulio C.; Scalabrin, Edson Emílio; Barthès, Jean-Paul

doi:10.1080/08839510701526954

Cited by 17 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The aim is to analyze the agent's performance when an optimal or near-optimal policy has been discovered, and to observe the agent's capacity then to adapt itself to a modified environment. Enembreck et al (Enembreck et al, 2007) have shown that this is a good way to observe the behavior of an adaptive agent. We have analyzed the agent's adaptation with the k-NR and Q-Learning algorithms.…”

Section: K-nr Evaluationmentioning

confidence: 99%

Unified Algorithm to Improve Reinforcement Learning in Dynamic Environments - An Instance-based Approach

Ribeiro

Favarim

Barbosa

et al. 2012

Proceedings of the 14th International Conference on Enterprise Information Systems

View full text Add to dashboard Cite

This paper presents an approach for speeding up the convergence of adaptive intelligent agents using reinforcement learning algorithms. Speeding up the learning of an intelligent agent is a complex task since the choice of inadequate updating techniques may cause delays in the learning process or even induce an unexpected acceleration that causes the agent to converge to a non-satisfactory policy. We have developed a technique for estimating policies which combines instance-based learning and reinforcement learning algorithms in Markovian environments. Experimental results in dynamic environments of different dimensions have shown that the proposed technique is able to speed up the convergence of the agents while achieving optimal action policies, avoiding problems of classical reinforcement learning approaches.

show abstract

Section: K-nr Evaluationmentioning

confidence: 99%

Unified Algorithm to Improve Reinforcement Learning in Dynamic Environments - An Instance-based Approach

Ribeiro

Favarim

Barbosa

et al. 2012

Proceedings of the 14th International Conference on Enterprise Information Systems

View full text Add to dashboard Cite

show abstract

“…Figure 2 shows the architecture of an adaptive trade agent proposed in Ref. 4. Upon receiving an offer/counter-offer, the agent classifies it as interesting in case it has been accepted by the corresponding peer and non-interesting otherwise.…”

Section: Ensemble-based Drift Detection For Drifting Negotiationmentioning

confidence: 99%

“…In this section we use the offer definitions proposed in Ref. 4. Table 1 presents each of the issues used in this work.…”

Section: Offer Spacementioning

confidence: 99%

Learning Negotiation Policies Using Ensemble-Based Drift Detection Techniques

Enembreck

Tacla

Barthès

2009

Int. J. Artif. Intell. Tools

Self Cite

View full text Add to dashboard Cite

International audienceIn this work we compare drift detection techniques and we show how they can improve the performance of trade agents in multi-issue bilateral dynamic negotiations. In a dynamic negotiation the utility values and functions of trade agents can change on the fly. Intelligent trade agents must identify and take such drift in the competitors into account changing also the offer policies to improve the global utility throughout the negotiation. However, traditional learning mechanisms disregard possible changes in a competitor's offer/counter-offer policy. In that case, the agent performance may decrease drastically. In our approach, a trade agent has a staff of weighted learner agents used to predict interesting offers. The staff uses the Dynamic Weighted Majority (DWM) algorithm to adapt itself creating, deleting and adapting staff members. The results obtained with the IB3 (Instance-based) learners and IB3-DWM learners show that ensemble methods like DWM are suitable for correctly identifying changes in agent negotiations

show abstract

“…The goal is to analyze the agent's performance when an optimal or near-optimal policy was discovered and observe the agent's capacity to adapt itself again to environments with other features (see Figure 7). Enembreck et al [6] shown that this is a good way to observe the behavior of an adaptive agent. We have analyzed the agent's adaptation with the Q ε -learning and Q-PR-BAP algorithms.…”

Section: A Performance Of the Agent With Q-pr-bapmentioning

confidence: 99%

A strategy for converging dynamic action policies

Ribeiro

Borges

Koerich

et al. 2009

2009 IEEE Symposium on Intelligent Agents

Self Cite

View full text Add to dashboard Cite

Abstract1 -In this paper we propose a novel strategy for converging dynamic policies generated by adaptive agents, which receive and accumulate rewards for their actions. The goal of the proposed strategy is to speed up the convergence of such agents to a good policy in dynamic environments. Since it is difficult to have the good value for a state due to the continuous changing in the environment, previous policies are kept in memory for reuse in future policies, avoiding delays or unexpected speedups in the agent's learning. Experimental results on dynamic environments with different policies have shown that the proposed strategy is able to speed up the convergence of the agent while achieving good action policies.

show abstract

Learning Drifting Negotiations

Cited by 17 publications

References 10 publications

Unified Algorithm to Improve Reinforcement Learning in Dynamic Environments - An Instance-based Approach

Unified Algorithm to Improve Reinforcement Learning in Dynamic Environments - An Instance-based Approach

Learning Negotiation Policies Using Ensemble-Based Drift Detection Techniques

A strategy for converging dynamic action policies

Contact Info

Product

Resources

About