Probabilistic reasoning in intelligent systems: networks of plausible inference
Probabilistic reasoning in intelligent systems: networks of plausible inference
Instance-Based Learning Algorithms
Machine Learning
Restructurable representations of negotiation
Management Science
The weighted majority algorithm
Information and Computation
Bayesian learning in negotiation
International Journal of Human-Computer Studies - Evolution and learning in multiagent systems
Dynamic Weighted Majority: A New Ensemble Method for Tracking Concept Drift
ICDM '03 Proceedings of the Third IEEE International Conference on Data Mining
Learning on opponent's preferences to make effective multi-issue negotiation trade-offs
ICEC '04 Proceedings of the 6th international conference on Electronic commerce
LEARNING DRIFTING NEGOTIATIONS
Applied Artificial Intelligence
Opponent modelling in automated multi-issue negotiation using Bayesian learning
Proceedings of the 7th international joint conference on Autonomous agents and multiagent systems - Volume 1
| Hi-index | 0.00 |
In an automated negotiation process, it is usual for one party not to reveal their preferences to its opponent. In an attempt to adjust itself to this process, each trade agent can be endowed with capabilities of learning and detecting its opponent's changing preferences. This paper presents techniques for drift detection that are useful in this scenario: the Instance-Based Learning algorithms (IB1, IB2 and IB3) and the Bayesian Networks. Theoretically, a group of expert agents is able to achieve better results than an individual, therefore, the DWM is part of these studies, once it provides an effective strategy for integrating the decision of several expert agents. The experiments performed revolve around the settings DWM-IB3 and DWM-Bayesian Network. The performance of each system was evaluated for gradual, moderate and abrubt drifts in concept, and the results showed that both settings are able to efficiently detect drifts in the preferences of the opponent.