Agents In Multi-agent Environment Research Articles

Methodology and Multicriteria Algorithm for Group Decision Support in Classification Problems

In this work, a Group Decision methodology and algorithm for small collaborating teams is introduced. It is based on a multicriteria algorithm for classification decisions, where aggregation of member preferences is executed at the parameter level. The algorithm applies to relatively well-structured problems guided by a process facilitator. Initially, a set of parameters is proposed by the facilitator to the group and next group members evaluate the proposed parameter set and express their preferences in numeric or linguistic format. Individual preferences are aggregated by appropriate operators, and a set of group parameter values is generated, which is used as input for the classification algorithm. NeXClass multicriteria classification algorithm is used for the classification of alternatives, initially at a training set of alternatives and later at the entire set. Finally, group members evaluate results, and consensus, as well as satisfaction metrics, are calculated. In case of a low acceptance level, problem parameters are reviewed by the facilitator, and the aggregation phase is repeated. The methodology is a valid approach for group decision problems and can be utilized in numerous business environments. The algorithm can be also utilized by software agents in multiagent environments for automated decision-making, given the large volume of agent-based decision-making in various settings today.

Herd’s Eye View: Improving Game AI Agent Learning with Collaborative Perception

We present a novel perception model named Herd's Eye View (HEV) that adopts a global perspective derived from multiple agents to boost the decision-making capabilities of reinforcement learning (RL) agents in multi-agent environments, specifically in the context of game AI. The HEV approach utilizes cooperative perception to empower RL agents with a global reasoning ability, enhancing their decision-making. We demonstrate the effectiveness of the HEV within simulated game environments and highlight its superior performance compared to traditional ego-centric perception models. This work contributes to cooperative perception and multi-agent reinforcement learning by offering a more realistic and efficient perspective for global coordination and decision-making within game environments. Moreover, our approach promotes broader AI applications beyond gaming by addressing constraints faced by AI in other fields such as robotics. The code is available at https://github.com/andrewnash/Herds-Eye-View

Developing, evaluating and scaling learning agents in multi-agent environments

The Game Theory & Multi-Agent team at DeepMind studies several aspects of multi-agent learning ranging from computing approximations to fundamental concepts in game theory to simulating social dilemmas in rich spatial environments and training 3-d humanoids in difficult team coordination tasks. A signature aim of our group is to use the resources and expertise made available to us at DeepMind in deep reinforcement learning to explore multi-agent systems in complex environments and use these benchmarks to advance our understanding. Here, we summarise the recent work of our team and present a taxonomy that we feel highlights many important open challenges in multi-agent research.

Negotiation Planning Of Autonomous Agents In Multi-agent Environment

In negotiation process, when a conflict arises between people in a society, each will try to pose some arguments to resolve the conflict and convince the other party. Successful termination of many of these interactions is not achieved only via exchange of a series of single-message suggestions and requires that an arguer offer compelling justifications to prove his/her claim. Obviously, before negotiation is started, decision should be made about the terms and arguments that can convince the other side and terminate negotiations successfully. In other words, an exact plan of arguments and their order should be prepared. The same is true about a multi - agent environment where each agent needs planning and executing a sequence of actions to achieve its goals. Some actions in these sequences are under the direct control of the planner agent and others require negotiations with other agents. Negotiation is usually carried out during the plan execution, however, it can be considered during the planning stage, as in real life. In this paper, we present a new approach that takes into account incomplete information about opponent’s position in negotiation by using a nondeterministic planner with incomplete knowlage and sensing actions. We evaluate our approach by solving a well known negotiation scenario.

Calculating the Reliability and Reputation of Agents in E-Commerce Multi-Agent Environments

E-commerce multi-agent systems have some features of autonomy and commercial transaction that improve rapidly and become a powerful tool for buying and selling products online. While multi-agent systems can break interactions in distributed and heterogeneous environments, malicious behaviors of the agents are a potential hazard in distributed e-commerce applications. Hence, calculating the reliability and reputation of agents play a critical role in determining how interactions occur and who to interact with. The way of calculating the reliability and reputation of each agent should be based on different elements, and there are no standard elements to compute reliability and reputation of each agent in multi-agent environment. This paper analyses some recent proposed methods in calculating reputation and reliability of each agent. The ultimate aim of this study is to define the standard elements which should be considered to calculate the reliability and reputation of agents in e-commerce multi-agent environment.