In this paper we study Relational Reinforcement Learning in a multi-agent setting. There is growing evidence in the Reinforcement Learning research community that a relational representation of the state space has many benefits over a propositional one. Complex tasks as planning or information retrieval on the web can be represented more naturally in relational form. Yet, this relational struct...

Learning to navigate in 3D environments from raw sensory input is an important step towards bridging the gap between human players and artificial intelligence in digital games. Recent advances in deep reinforcement learning have seen success in teaching agents to play Atari 2600 games from raw pixel information where the environment is always fully observable by the agent. This is not true for ...

Simulations of artificial stock markets were considered as early as 1964 [Sti64] and multi-agent ones were introduced as early as 1989 [KM89]. Starting the early 90's [PAH + 94, LLS00, Ter00], collaborations of economists and physicists produced increasingly realistic simulation platforms. Currently, the market stylized facts are easily reproduced and one has now to address the realistic detail...

We have constructed models for a series of platinum-DNA adducts that represent the binding of two agents, [Pt2(DTBPA)Cl2](II) and [Pt2(TPXA)Cl2](II), to DNA via inter- and intra-strand cross-linking, and carried out molecular dynamics simulations and DNA conformational dynamics calculations. The effects of trans- and cis-configurations of the centers of these di-nuclear platinum agents, and of ...

We propose an infrastructure for the design and development of Open Interaction Systems (OISs), based on solutions from the areas of Service Oriented Architecture, Semantic Technologies and Normative Multiagent Systems, in particular the OCeAN metamodel of Artificial Institution. OISs are open to diverse types of participants (i.e., software agents), and enable them to interact with each other ...

Agent-based models are important tools for simulating farmers’ behaviour in response to changing environmental, economic or institutional conditions and policies. This article introduces an agent-based modelling approach that combines behavioural factors with standard bio-economic of agricultural production. More specifically, our framework integrates the cumulative prospect theory social inter...

The exploration of social networks is essential to find capable collaborators who can help problem-solving and to augment cooperation between users. This paper describes an agent based social networking system called PeCo-Mediator-II for seeking for a collaborator with the chain of personal connections (PeCo) in distributed organizations. This system helps gathering, exploring, and visualizing ...

agent technology is an emerging and promising research area in software technology, which increasingly contributes to the development of value-added information systems for large healthcare organizations due to its capability to automatically modify themselves in response to changes in their operating environment. multi agent system (mas) approach is the construction of a complex system as a se...

We present a concept for facilitating the development of agent applications called agent usage patterns. These patterns are meant to bridge the conceptual gap that currently exists between the programmer and agent-based middleware systems. We analyze each pattern's potential for improving performance, reliability, and security. Finally, we describe some preliminary experiments: we present a sto...