Catálogo de publicaciones - libros

We consider the problem of designing error correcting codes (ECC), a hard combinatorial optimization problem of relevance in the field of telecommunications. This problem is firstly approached via a battery of local search (LS) methods that are compared and analyzed. Then, we study how to tackle this problem by having a society of interacting autonomous agents where each agent is endowed with a specific (not necessarily unique) strategy for local improvement. Distinct topologies and forms of interaction are analyzed and discussed in the paper. Specifically, it is shown how the election of the LS methods and their combination influences the results. An empirical evaluation shows that agent-based models are promising to solve of this problem.

Reinforcement learning (RL) agents can benefit from adaptive exploration/exploitation behavior, especially in dynamic environments. We focus on regulating this exploration/exploitation behavior by controlling the action-selection mechanism of RL. Inspired by psychological studies which show that affect influences human decision making, we use artificial affect to influence an agent’s action-selection. Two existing affective strategies are implemented and, in addition, a new hybrid method that combines both. These strategies are tested on ‘maze tasks’ in which a RL agent has to find food (rewarded location) in a maze. We use Soar-RL, the new RL-enabled version of Soar, as a model environment. One task tests the ability to quickly adapt to an environmental change, while the other tests the ability to escape a local optimum in order to find the global optimum. We show that artificial affect-controlled action-selection in some cases helps agents to faster adapt to changes in the environment.

This paper presents an architecture of an agent-based decision support system (ADSS) for ecological-medical situation assessment. The system receives statistical information in form of direct and indirect pollution indicator values. The ultimate goal of the modeled multi-agent system (MAS) is to evaluate the impact of the exposure to pollutants in population health. The proposed ADSS interacts with humans in real-time “what-if” scenarios, providing the user with evidence for optimal decision making. A detailed description of all the agents and their BDI (beliefs, desires, intentions) cards is presented.

The paper introduces an approach to using a meta-ontology framework for complex multi-agent systems design, and illustrates it in an application related to ecological-medical issues. The described shared ontology is pooled from private sub-ontologies, which represent a problem area ontology, an agent ontology, a task ontology, an ontology of interactions, and the multi-agent system architecture ontology.

This work presents the experience on designing a multiagent system devoted to the transportation of passengers using the PASSI methodology. The agent system is in charge of the planning and scheduling of passenger trips using the contract-net protocol as base coordination mechanism. It also allows the events’ processing caused by changes to the original plan due to vehicle failures or delays, detours and traffic jams, cancellations or passenger no-show. The system has been modeled with the PASSI Toolkit (PTK) and implemented over Jade.

The use of surveillance systems has grown exponentially during the last decade. Moreover, the agency paradigm has shown to be suitable for the design and development of complex systems such as surveillance systems. They provide autonomy, reactivity, social ability and pro-activeness to carry out surveillance tasks in a semi-automatic way, collaborating with users in a more effective manner. Agents provide coordination mechanisms, solve conflicts, and determine through negotiation processes the more appropriate distribution of the surveillance tasks. Existent agent-based surveillance systems do not really use agent-based methodologies to develop them. In this paper, our experience for modelling advanced surveillance systems using the INGENIAS methodology is described.

In this paper we present , a novel dynamic Web tool for managing probabilistic and possibilistic temporal constraint networks. These networks are a special sort of temporal constraint satisfaction problems, useful for representing and reasoning with uncertain temporal relations between temporal points. They can be applied as an effective formalism in a very different kind of domains. has been developed using dynamic Web technology for mainly, make easier the complex process of interpreting, evaluating, and finally, extracting useful knowledge from the network. It can be viewed as a novel knowledge acquisition tool and, for practical purposes, it will be used for work with temporal patterns extracted from temporal data mining processes.

Biomedical research is now information intensive; the volume and diversity of new data sources challenges current database technologies. The development and tuning of database technologies for biology and medicine will maintain and accelerate the current pace for innovation and discovery. New promising application fields such as the Semantic Web and Semantic Web Services can leverage the potential of biomedical information integration and discovery, facing the problem of semantic heterogeneity of biomedical information sources in a variety of storage and data formats widely distributed both across the Internet and within individual organizations. In this paper, we present BIRD, a fully-fledged biomedical information integration solution that combines natural language analysis and semantically-empowered techniques to ascertain how the user needs can be best fit. Our approach is backed with a proof-of-concept implementation where the breakthrough and efficiency of integrating the biomedical publications database PubMed, the Database of Interacting Proteins (DIP) and the Munich Information Center for Protein Sequences (MIPS) has been tested.

This paper depicts an already developed experience in search for a predictable mechanism with respect to the future performance of a student considering the numerous factors that influence in its failure/success. The use of different neural networks configurations in conjunction with a large data volume on top of detailed attributes consideration for each student makes for an adequate base for the results obtained to be analyzed. The idea behind this paper is to arrange a mechanism that allows us to estimate before hand taking into consideration data from the student in reference to family, social and wealth surroundings for the student future performance identifying those factors that favors the tendency to failure or success.

In this work we present power efficiency optimization for microwave ovens by means of genetic algorithms (GA). Two kind of microwave applicators are analyzed in this case: cylindrical and rectangular ones. In the first case, optimization of the oven uses cavity dimensions, waveguide feeding location and polarization as design parameters. In the second case, waveguide slots are used to feed the rectangular multimode cavity. The slots’ dimension, position and angle in the waveguide are optimized by the genetic algorithm in order to achieve the best power efficiency. All simulations are carried out by the CST Microwave Studio, a Finite Integration Technique (FIT) commercial software capable of solving electromagnetic (EM) structures in three dimensions whereas GA are implemented in Matlab. The obtained results show that combination of both the EM software and GA provides a very powerful tool for microwave oven design and optimization.