Tesis sobre el tema "Artificial adaptive agents"

The work presented in this document is part of the project AutoFocus: Adaptive Self-Improving Multi-Agent Systems' that is being developed at the research unit LabMAg, which objective is the implementation of multi-agent systems based on autonomous entities capable of self-optimized and adaptive behaviors. The notion of autonomic computation, like other notions that also imply pro-active computation, is based on autonomous entities that actively work to achieve their objectives and have the ability to dynamically adjust to changes in their environment, constrained by time and resource limits. In the approach used by the AutoFocus project, that adaptation to change and the regulation of the agent's capabilities, result from the combination of cognitive aspects with emotional based aspects. The agent model defined and used by the AutoFocus project is the Agent Flow Model. The task that corresponded to the work presented in this document was to develop a platform for the Agent Flow Model. It was intended, with this platform, to provide a tool that enables the rapid deployment and monitoring of agents based on this model. The developed work consisted in the analysis and design, oriented to objects, implementation and testing of components of this platform

This thesis reports the development and study of novel architectures for the simulation of adaptive behaviour based on artificial neural networks. There are two distinct levels of enquiry. At the primary level, the initial aim was to design and implement a unified architecture integrating sensorimotor learning and overall control. This was intended to overcome shortcomings of typical behaviour-based approaches in reactive control settings. It was achieved in two stages. Initially, feedforward neural networks were used at the sensorimotor level of a modular architecture and overall control was provided by an algorithm. The algorithm was then replaced by a recurrent neural network. For training, a form of reinforcement learning was used. This posed an intriguing composite of the well-known action selection and credit assignment problems. The solution was demonstrated in two sets of simulation studies involving variants of each architecture. These studies also showed: firstly that the expected advantages over the standard behaviour-based approach were realised, and secondly that the new integrated architecture preserved these advantages, with the added value of a unified control approach. The secondary level of enquiry addressed the more foundational question of whether the choice of processing mechanism is critical if the simulation of adaptive behaviour is to progress much beyond the reactive stage in more than a trivial sense. It proceeded by way of a critique of the standard behaviourbased approach to make a positive assessment of the potential for recurrent neural networks to fill such a role. The findings were used to inform further investigations at the primary level of enquiry. These were based on a framework for the simulation of delayed response learning using supervised learning techniques. A further new architecture, based on a second-order recurrent neural network, was designed for this set of studies. It was then compared with existing architectures. Some interesting results are presented to indicate the appropriateness of the design and the potential of the approach, though limitations in the long run are not discounted.

The modern trend in electricity industries around the world is towards privatisation. Increased competition, it is argued, will ultimately benefit the consumer. However, the particular nature of electricity generation and supply means strong regulation of a privatised market will always be necessary. In establishing a privatised industry, decisions need to be made about the mechanisms governing the requirements to meet demand, to maintain the viability of the network and to ensure generators are paid correctly for power generated. Unfortunately, it is unclear what processes to use to achieve these goals while still delivering some benefit to the consumer in the form of reduced electricity costs. This research, sponsored by the National Grid Company, examines whether the application of new ideas in artificial intelligence could offer the potential for gaining insights into the affects of certain market mechanisms on the competitors in the market. Our approach to gaining greater understanding into how the market operates is to adopt an evolutionary economics perspective. We have constructed autonomous adaptive agents to represent the generating companies in a simplified model of the UK market in electricity generation. The main body of the thesis contains a description of the process of developing the model and the agent architecture. Once we were satisfied that the model incorporated some key features of the real world market and that the agents, based on learning classifier systems, were able to perform well in simpler environments, we examined how multiple adaptive agents learn to interact in the simplified model. We conclude that the agents are able to learn how to behave in ways analogous to the observed behaviour of real world generating companies. We then illustrate the potential for this type of economic model by examining how alterations to market structure affect agent behaviour, and investigate to what extent the agents are able to learn how to cooperate for mutual long term benefit.

Managing the information flow within a big organization is a challenging task. Moreover, in a distributed decision-making process conflicting objectives occur. In this paper, artificial adaptive agents are used to analyze this problem. The decision makers are implemented as Classifier Systems, and their learning process is simulated by Genetic Algorithms. To validate the outcomes we compared the results with the optimal solutions obtained by full enumeration. It turned out that the genetic algorithm indeed was able to generate useful rules that describe how the decision makers involved in new product development should react to the requests they are required to fulfill. (author's abstract)
Series: Working Papers SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

Thesis (M.S. in Information Technology Management and M.S. in Modeling Virtual Environment and Simulation)--Naval Postgraduate School, March 2004.
Thesis advisor(s): Alex Bordetsky, John Hiles. Includes bibliographical references. Also available online.

The attempt to evolve complete embodied and situated artificial creatures in which both morphological and control characteristics are adapted during the evolutionary process has been and still represents a long term goal key for the artificial life and the evolutionary robotics community. Loosely inspired by ancient biological organisms which are not provided with a central nervous system and by simple organisms such as stick insects, this thesis proposes a new genotype encoding which allows development and evolution of mor- phology and neural controller in artificial agents provided with a distributed neural network. In order to understand if this kind of network is appropriate for the evolution of non trivial behaviours in artificial agents, two experiments (description and results will be shown in chapter 3) in which evolution was applied only to the controller’s parameters were performed. The results obtained in the first experiment demonstrated how distributed neural networks can achieve a good level of organization by synchronizing the output of oscillatory elements exploiting acceleration/deceleration mechanisms based on local interactions. In the second experiment few variants on the topology of neural architecture were introduced. Results showed how this new control system was able to coordinate the legs of a simulated hexapod robot on two different gaits on the basis of the external circumstances. After this preliminary and successful investigation, a new genotype encoding able to develop and evolve artificial agents with no fixed morphology and with a distributed neural controller was proposed. A second set of experiments was thus performed and the results obtained confirmed both the effectiveness of genotype encoding and the ability of distributed neural network to perform the given task. The results have also shown the strength of genotype both in generating a wide range of different morphological structures and in favouring a direct co-adaptation between neural controller and morphology during the evolutionary process. Furthermore the simplicity of the proposed model has showed the effective role of specific elements in evolutionary experiments. In particular it has demonstrated the importance of the environment and its complexity in evolving non-trivial behaviours and also how adding an independent component to the fitness function could help the evolutionary process exploring a larger space solutions avoiding a premature convergence towards suboptimal solutions.

An intelligent agent must be capable of using its past experience to develop an understanding of how its actions affect the world in which it is situated. Given some objective, the agent must be able to effectively use its understanding of the world to produce a plan that is robust to the uncertainty present in the world. This thesis presents a novel computational framework called the Adaptive Modelling and Planning System (AMPS) that aims to meet these requirements for intelligence. The challenge of the agent is to use its experience in the world to generate a model. In problems with large state and action spaces, the agent can generalise from limited experience by grouping together similar states and actions, effectively partitioning the state and action spaces into finite sets of regions. This process is called abstraction. Several different abstraction approaches have been proposed in the literature, but the existing algorithms have many limitations. They generally only increase resolution, require a large amount of data before changing the abstraction, do not generalise over actions, and are computationally expensive. AMPS aims to solve these problems using a new kind of approach. AMPS splits and merges existing regions in its abstraction according to a set of heuristics. The system introduces splits using a mechanism related to supervised learning and is defined in a general way, allowing AMPS to leverage a wide variety of representations. The system merges existing regions when an analysis of the current plan indicates that doing so could be useful. Because several different regions may require revision at any given time, AMPS prioritises revision to best utilise whatever computational resources are available. Changes in the abstraction lead to changes in the model, requiring changes to the plan. AMPS prioritises the planning process, and when the agent has time, it replans in high-priority regions. This thesis demonstrates the flexibility and strength of this approach in learning intelligent behaviour from limited experience.

In this work, I view incremental experiential learning in intelligent software agents as progressive agent self-adaptation. When an agent produces an incorrect behavior, then it may reflect on, and thus diagnose and repair, the reasoning and knowledge that produced the incorrect behavior. In particular, I focus on the self-diagnosis and self-repair of an agent's domain knowledge. The implementation of systems with the capability to self-diagnose and self-repair involves building both reasoning processes capable of such learning and knowledge representations capable of supporting those reasoning processes. The core issue my dissertation addresses is: what kind of metaknowledge (knowledge about knowledge) may enable the agent to diagnose faults in its domain knowledge? In providing a solution to this issue, the central contribution of this research is a theory of the kind of metaknowledge that enables a system to reason about and adapt its conceptual knowledge. For this purpose, I propose a representation that explicitly encodes metaknowledge in the form of procedures called Empirical Verification Procedures (EVPs). In the proposed knowledge representation, an EVP is associated with each concept within the agent's domain knowledge. Each EVP explicitly semantically grounds the associated concept in the agent's perception, and can thus be used as a test to determine the validity of knowledge of that concept during diagnosis. I present the formal and empirical evaluation of a system, Augur, that makes use of EVP metaknowledge to adapt its own domain knowledge in the context of a particular subclass of classification problem that I call compositional classification, in which the overall classification task can be broken into a hierarchically organized set of subtasks. I hypothesize that EVP metaknowledge will enable a system to automatically adapt its knowledge in two ways: first, by adjusting the ways that inputs are categorized by a concept, in accordance with semantics fixed by an associated EVP; and second, by adjusting the semantics of concepts themselves when they fail to contribute appropriately to system goals. The latter adaptation is realized by altering the EVP associated with the concept in question. I further hypothesize that the semantic grounding of domain concepts in perception through the use of EVPs will increase the generalization power of a learner that operates over those concepts, and thus make learning more efficient. Beyond the support of these hypotheses, I also present results pertinent to the understanding of learning in compositional classification settings using structured knowledge representations.

This master thesis investigates the principal capability of artificial evolution to produce tool use behavior in adaptive agents, excluding the application of life-time learning or adaptation mechanisms. Tool use is one aspect of complex behavior that is expected from autonomous agents acting in real-world environments. In order to achieve tool use behavior an agent needs to identify environmental objects as potential tools before it can use the tools in a problem-solving task. Up to now research in robotics has focused on life-time learning mechanisms in order to achieve this. However, these techniques impose great demands on resources, e.g. in terms of memory or computational power. All of them have shown limited results with respect to a general adaptivity. One might argue that even nature does not present any kind of omni-adaptive agent. While humans seem to be a good example of natural agents that master an impressive variety of life conditions and environments (at least from a human perspective, other examples are spectacular survivability observations of octopuses, scorpions or various viruses) even the most advanced engineering approaches can hardly compete with the simplest life-forms in terms of adaptation. This thesis tries to contribute to engineering approaches by promoting the application of artificial evolution as a complementing element with the presentation of successful pioneering experiments. The results of these experiments show that artificial evolution is indeed capable to render tool use behavior at different levels of complexity and shows that the application of artificial evolution might be a good complement to life-time approaches in order to create agents that are able to implicitly extract concepts and display tool use behavior. The author believes that off-loading at least parts of the concept retrieval process to artificial evolution will reduce resource efforts at life-time when creating autonomous agents with complex behavior such as tool use. This might be a first step towards the vision of a higher level of autonomy and adaptability. Moreover, it shows the demand for an experimental verification of commonly accepted limits between qualities of learned and evolved tool use capabilities.

This thesis is concerned with the analysis of economic network formation. There are three novel sections to this thesis (Chapters 5, 6 and 8). In the first, the non-cooperative communication network formation model of Bala and Goyal (2000) (BG) is re-assessed under conditions of no inertia. It is found that the Strict Nash circle (or wheel) structure is still the equilibrium outcome for n = 3 under no inertia. However, a counter-example for n = 4 shows that with no inertia infinite cycles are possible, and hence the system does not converge. In fact, cycles are found to quickly dominate outcomes for n > 4 and further numerical simulations of conditions approximating no inertia (probability of updating > 0.8 to 1) indicate that cycles account for a dramatic slowing of convergence times. These results, together with the experimental evidence of Falk and Kosfeld (2003) (FK) motivate the second contribution of this thesis. A novel artificial agent model is constructed that allows for a vast strategy space (including the Best Response) and permits agents to learn from each other as was indicated by the FK results. After calibration, this model replicates many of the FK experimental results and finds that an externality exploiting ratio of benefits and costs (rather than the difference) combined with a simple altruism score is a good proxy for the human objective function. Furthermore, the inequity aversion results of FK are found to arise as an emergent property of the system. The third novel section of this thesis turns to the nature of network formation in a trust-based context. A modified Iterated Prisoners' Dilemma (IPD) model is developed which enables agents to play an additional and costly network forming action. Initially, canonical analytical results are obtained despite this modification under uniform (non-local) interactions. However, as agent network decisions are 'turned on' persistent cooperation is observed. Furthermore, in contrast to the vast majority of non-local, or static network models in the literature, it is found that a-periodic, complex dynamics result for the system in the long-run. Subsequent analysis of this regime indicates that the network dynamics have fingerprints of self-organized criticality (SOC). Whilst evidence for SOC is found in many physical systems, such dynamics have been seldom, if ever, reported in the strategic interaction literature.

Le CERV constitue à Brest un pôle d'excellence en réalité virtuelle à vocation européenne. Les systèmes que l'on cherche à modéliser sont de plus en plus complexes, du fait essentiellement de la diversité des composants, de la diversité des structures et de la diversité des interactions mises en jeu. Un système est alors a priori un milieu ouvert (apparition/disparition dynamique de composants) et hétérogène (morphologies et comportements variés). La réalité virtuelle implique pleinement l'utilisateur humain dans la simulation, rejoignant ainsi l'approche de la conception. La majorité des travaux en réalité virtuelle concerne l'immersion sensorimotrice de l'utilisateur humain au sein d'univers. Ces univers virtuels offrent à l'utilisateur la sensation d'être dans l'environnement et lui donne la possibilité d'y agir. Pour être complet, il faut également "qu'il s'y passe quelque chose", et pas seulement en terme de résultat des actions de l'utilisateur. Les entités qui peuplent les univers virtuels doivent donc avoir un comportement autonome. Ceci soulève la question suivante : comment doter une entité d'un comportement autonome dans un environnement virtuel complexe auquel l'homme participe ? Des techniques d'intelligence artificielle symbolique ont déjà été appliquées pour définir ces comportements. Mais ces techniques montrent très vite leurs limites car elles sont principalement basées sur des règles de comportements mises a priori par le concepteur. Or, dans des mondes virtuels complexes (simulation ouverte, hétérogène et participative), plusieurs entités vont avoir des comportements imprédictibles (variabilité comportementale des entités autonomes, libre arbitre de l'utilisateur), créant ainsi des situations toujours nouvelles. Et face à une situation non prévue par le programmeur, les entités auront le plus souvent des comportements inadaptés. C'est pourquoi les méthodologies tirées des systèmes artificiels adaptatifs peuvent contribuer à pallier ces limitations. Le travail que je poursuis porte sur la thématique de l'adaptation de comportements d'entités autonomes en environnement virtuel participatif. Adapter son comportement, c'est effectuer des transformations conduisant à s'adapter à son environnement. Cette adaptation aura pour objectif de rendre le comportement de l'entité virtuelle le plus crédible possible (ressemblant à un comportement humain). Pour cela, nous prenons le parti de considérer que l'entité doit apprendre au fur et à mesure des expériences, elle doit anticiper le comportement des autres entités et les conséquences sur l'environnement, elle doit également exploiter la présence de l'utilisateur humain dans l'univers virtuel pour adapter son comportement. Imaginons un monde virtuel où chaque entité, au même titre qu'un humain, aurait son propre comportement qui évolue automatiquement pendant la simulation. C'est tout l'enjeu des travaux de recherche présentés ici.

As a consequence of global environmental change, sustainable management and governance of natural resources face critical challenges, such as dealing with non-linear dynamics, increased resource variability, and uncertainty. This thesis seeks to address some of these challenges by using simulation models. The first line of research focuses on the use of learning-by-doing (LBD) for managing a renewable resource, exemplified by a fish stock, and explores LBD in a theoretical model using artificial intelligence (Paper I and II). The second line of research focuses on the emergence of different forms of self-governance and their interrelation with the dynamics of trust among fishers when harvesting a shared resource, using an agent-based model. This model is informed by qualitative data based on small-scale fisheries in Mexico (Paper III and IV). Paper I and II find that the most sustainable harvesting strategy requires that the actor values current and future yields equally, cautiously experiments around what is perceived as the best harvest action, and rapidly updates its ‘mental model’ to any perceived change in catch. More specifically, Paper II reveals that understanding these aspects in relation to the type of change can yield not only increased performance, but also, and more importantly, increased robustness to both fast and slow changes in resource dynamics. However, when resource dynamics include the possibility of a more fundamental shift in system characteristics (a regime shift), LBD is problematic due to the potential for crossing a threshold, resulting in possible persistent reductions in harvests (Paper I). In Paper III, results indicate that cooperative forms of self-governance are more likely to establish and persist in communities where fishers’ have prior cooperative experience, fishers’ trustworthiness is more or less equal, and that this likelihood increases when resource availability fluctuates seasonally. Finally, to achieve a transformation toward more cooperative forms of self-governance, interventions are required that can strengthen both financial capital and trust among the members of the cooperatives (Paper IV). The unique contribution of this thesis lies in the method for ‘quantitatively’ studying LBD, the stylized model of a small-scale fishery, and the analysis of the two models to advance our understanding of processes of learning and self-governance in uncertain and variable social-ecological environments. Together, the results shed light on how social and ecological factors and processes co-evolve to shape social-ecological outcomes, as well as contributing to the development of novel methods within the emerging field of sustainability science.
I vårt antropocena tidevarv är ett långsiktigt förvaltarskap av naturresurser inom social-ekologiska system av yttersta vikt. Detta kräver en djup förståelse av människan, ekologin, interaktionerna sinsemellan och deras utveckling över tid. Syftet med denna avhandling är att nå en djupare och mer nyanserad förståelse kring två av grundpelarna inom forskningen av hållbar förvaltning av naturresurser–kontinuerligt lärande genom learning-by-doing (LBD) för att förstå naturresursens dynamik, samt vad som kan kallas socialt kapital, i detta sammanhang i betydelsen tillit mellan individer, som naturligtvis ligger till grund för framgångsrik gemensam förvaltning. Denna föresats operationaliseras genom att använda två olika simuleringsmodeller. Den ena modellen undersöker hur en hållbar förvaltning av en förnyelsebar resurs, i denna avhandling exemplifierad av en fiskepopulation, kan uppnås genom LBD. Den andra modellen söker blottlägga det komplexa sociala samspel som krävs för att praktisera gemensam förvaltning genom att använda ett fiskesamhälle som fallstudie. Tidigare forskning på båda dessa två områden är relativt omfattade. Emellertid har den forskning som specialiserat sig på LBD i huvudsak inskränkt sig till empiriska fallstudier. Vad som bryter ny mark i denna avhandling är att vi konstruerar en simuleringsmodell av LBD där vi kan studera lärandeprocessen i detalj för att uppnå en mer hållbar förvaltning över tid. Beträffande modellen som behandlar socialt kapital så har tidigare forskning fokuserat på hur en organisation, eller grupp, kan uppnå hållbar förvaltning. Dock saknas ett helhetsgrepp där som tar hänsyn till alla nivåer; från individnivå (mikro), via gruppnivå (meso), till samhällsnivå (makro). Detta är något som denna avhandling försöker avhjälpa genom att undersöka betydelsen av individers egenskaper, uppbyggnaden av socialt kapital, samt hur detta påverkar emergens av ett samhälle dominerat av mer kooperativa förvaltningsformer respektive mer hierarkiska diton. I papper I and II studeras kärnan av LBD som återkoppling mellan en aktör och en resurs, där aktören lär sig genom upprepade interaktioner med en resurs.  Resultaten visar att LBD är av avgörande betydelse för en hållbar förvaltning, speciellt då naturresursens dynamik är stadd i förändring. I den mest hållbara strategin bör aktören värdera nuvarande och framtida fångster lika högt, försiktigt experimentera kring vad aktören upplever som bästa strategi, för att sedan anpassa sin mentala modell till upplevda förändringar i fångst relativt dess insats någorlunda kraftigt. I papper III och IV behandlas uppbyggnaden av förtroende mellan individer och grupp, samt själv-organiserat styre. Genom att använda småskaligt fiske i Mexiko som en illustrativ fallstudie, utvecklades en agent-baserad modell av ett arketypiskt småskaligt fiskesamhälle. Resultaten indikerar att kooperativa förvaltningsformer är mer dominanta i samhällen där de som utför fisket har liknande pålitlighet, starkt gemensamt socialt kapital vid kooperativets start, och då resursen fluktuerar säsongsmässigt (papper III). Papper IV visar att för att uppnå en transformation från hierarkiska förvaltningsformer till kooperativa diton krävs interventioner som inriktar sig på både socialt och finansiellt kapital. Denna avhandling bidrar således till en djupare förståelse kring hur socialt kapital växer fram, samt hur mer strategiska LBD processer bör utformas när abrupta och osäkra förändringar i ekosystemen blir allt vanligare på grund av människans ökade tryck på planeten.

At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 2: Submitted. Paper 3: Submitted. Paper 4: Manuscript.

For underwater robotics there exists severe constraints regarding wireless bandwidth in the kilobits range. This makes a centralised approach to high-level mission management possibly less than ideal due to inherent delays and possible temporary incompleteness in data during decision making. This thesis aims to propose, implement (in ROS) and test a distributed approach. An auction based method for task assignment was being used, as well as a Pagerank based approach that models a trust based hierarchy between autonomous agents inferred from information exchange, in order to enforce decision conformity. Simulations where carried out using UWsim and a custom made bandwidth limiter for ROS. It was concluded that the Pagerank based algorithm managed to uphold conformity and solve conflicts during network slowdown but did not always lead to the correct decisions being enforced.

Les agents situés dans l'image fournissent un cadre privilégié pour la mise en oeuvre de stratégies coopératives et localement adaptées en segmentation d'image. Ils facilitent l'intégration des connaissances a priori, expressions d'un modèle, permettant ainsi de dégager de nouvelles contraintes indispensables à toutes les étapes de la vision par ordinateur (de la segmentation à l'interprétation). Nous proposons un cadre conceptuel pour l'architecture logicielle d'un système de vision bas-niveau basée sur des agents situés dans l'image. Une telle architecture est articulée en trois niveaux d'analyse et de description : 1. Description globale et structurelle de l'organisation regroupant les agents. Cette étape de description s'attache à établir les liens entre agents. Nous proposons comme élément organisationnel la pyramide irrégulière qui impose sa structure à la population d'agents, afin de garantir un comportement globalement contrôlable et convergent de ces derniers. 2. Description locale, fonctionnelle et comportementale des agents composant le système. Nous proposons une mise en oeuvre particulière de l'architecture logicielle de vision bas-niveau. Dans cette dernière, deux familles d'agents, qui traduisent des primitives région et contour, interagissent localement au sein de la pyramide. Notre objectif est de montrer comment cette méthodologie permet une implémentation riche, flexible et distribuée des aspects précédemment identifiés; à savoir : l'adaptation locale, l'intégration et l'expression d'incertitudes dans l'information a priori et des traitements coopératifs région/région et région/contour. 3. Finalement, une analyse globale, comparative et fonctionnelle vérifie que l'ensemble des interactions locales produit une bonne segmentation des images. Nous comparons notre approche avec d'autres méthodes de segmentation sur des images médicales et des images de synthèse.

Etablir le meilleur plan pour l'usinage d'un produit, le meilleur ordonnancement des activités de construction d'un bâtiment ou la meilleure tournée de véhicules pour la livraison des commandes, en prenant en compte diverses contraintes économiques, temporelles, humaines, ou même météorologiques : dans cette diversité d'applications, optimiser la planification est une tâche complexe par le grand nombre d'entités hétérogènes en interaction, la forte dynamique, les objectifs contradictoires à atteindre, etc. La planification de missions pour des constellations de satellites en est un exemple majeur : beaucoup de paramètres et de contraintes, souvent antagonistes, doivent être pris en compte, entraînant une importante combinatoire. Actuellement, en Europe, les plans de missions sont élaborés au sol, juste avant que le satellite ne soit visible par la station d'émission. Les requêtes arrivant durant la planification ne peuvent être traitées, et sont mises en attente. De plus, la complexité de ce problème croit drastiquement : le nombre de constellations et les satellites les composant augmentent, ainsi que le nombre de requêtes journalières. Les approches actuelles montrent leurs limites. Pour pallier à ces inconvénients, de nouveaux systèmes basés sur la décentralisation et la distribution inhérentes à ce genre de problèmes, sont nécessaires. La théorie des systèmes multi-agents adaptatifs (AMAS) et notamment le modèle AMAS4Opt (AMAS for Optimisation) ont montré leur adéquation pour la résolution de problèmes d'optimisation complexes sous contraintes. Le comportement local et coopératif des agents AMAS permet au système de s'auto-adapter à la forte dynamique et de fournir des solutions adéquates rapidement. Dans cette thèse, nous adressons la résolution de la planification des missions de satellites par AMAS. Pour cela, nous avons complété et enrichi les modèles d'agents proposés par AMAS4Opt. Nous avons ainsi développé le système de planification dynamique de missions ATLAS. Pour valider ATLAS sur divers critères, nous avons utilisé un grand nombre de données hétérogènes. Enfin, ce travail a été comparé à un système " opérationnel' " standard sur des scénarios réels, mettant en valeur les apports de notre système
Building the best plan in product treatment, the best schedule to a building construction or the best route for a salesman in order to visit a maximum of cities in the time allowed while taking into account different constraints (economic, temporal, humans or meteorological ): in all of those variety of applications, optimizing the planning is a complex task including a huge number of heterogeneous entities in interaction, the strong dynamics, multiple contradictory objectives, etc. Mission planning for constellations of satellites is a major example: a lot of parameters and constraints, often antagonists must be integrated, leading to an important combinatorial search space. Currently, in Europe, plans are built on ground, just before the satellite is visible by the ground stations. Any request coming during the planning process must wait for the next period. Moreover, the complexity of this problem grows drastically: the number of constellations and satellites increases, as the number of daily requests. Current approaches have shown their limits. To overcome those drawbacks, new systems based on decentralization and distribution inherent to this problem, are needed. The adaptive multi-agent systems (AMAS) theory and especially the AMAS4Opt (AMAS For Optimization) model have shown their adequacy in complex optimization problems solving. The local and cooperative behavior of agents allows the system to self-adapt to highly dynamic environments and to quickly deliver adequate solutions. In this thesis, we focus on solving mission planning for satellite constellations using AMAS. Thus, we propose several enhancement for the agent models proposed by AMAS4Opt. Then, we design the ATLAS dynamic mission planning system. To validate ATLAS on several criteria, we rely on huge sets of heterogeneous data. Finally, this work is compared to an operational and standard system on real scenarios, highlighting the value of our system

Cette thèse propose une nouvelle approche de la vision par ordinateur, insistant sur le rôle joué par les différentes informations et connaissances présentes dans un système intégré de vision. Ces connaissances peuvent être vues comme un but à atteindre dans la chaine de traitement ou comme un instrument pour guider ces différents traitement. Ce double rôle est mis en évidence en insistant sur la gestion des informations disponibles à tout moment pour effectuer la tâche de reconnaissance. Une approche distribuée est proposée, sous la forme d'un système multi-agents, où chaque agent est spécialisé pour la reconnaissance d'un concept de l'image. Le modèle générique d'agent est composé de quatre comportements de base : perception (croissance de région, suivi de contour), interaction (fusion de primitives), différenciation (interprétation de primitive) et reproduction (stratégie de focalisation des agents). Ces comportements prennent des décisions à l'aide d'évaluations multi-critères, faisant référence à plusieurs critères spécialisés. Des capacités d'adaptation et de coopération sont nécessaires aux agents pour mener à bien leurs tâches, et ce tant au niveau local qu'au niveau des groupes d'agents (travaillant sur un même groupement perceptuel). Ce système est appliqué pour la reconnaissance des cellules en mouvement. Chaque cellule est divisée en plusieurs composantes (noyau, pseudopode, cytoplasme, ...) qui servent de concepts de base pour la spécialisation des agents.

O Dilema do Prisioneiro com n jogadores é um problema que ilustra a dificuldade na formação da cooperação em sociedades de indivíduos racionais. Diversos trabalhos foram feitos no sentido de compreender melhor os fatores que influenciam o surgimento e a evolução da cooperação nessas sociedades, sendo que muitos desses mostraram que a simulação deste tipo de problema carece de escalabilidade, o que impede a realização de experimentos que envolvam uma grande quantidade de agentes ou de parâmetros de teste. Este trabalho tem o intuito de aplicar conceitos de computação paralela para tratar este problema. Para tal, foi desenvolvido um sistema denominado PS2 E2 , evolução de um trabalho anterior, cuja utilização em alguns cenários possibilitou a verificação da influência de alguns parâmetros tais como o tamanho da população e a expressividade do modelo de representação de estratégias na utilidade global de um conjunto de agentes que jogam o Dilema do Prisioneiro com n jogadores.
The n-Player Prisoners Dilemma is a problem that illustrates the difficulty of cooperation formation in societies composed of rational individuals. Several studies were made to better understand the factors that influence the emergence and evolution of cooperation in these societies. Many of these showed that the simulation of this type of problem lacks scalability, which hinders the achievement of experiments involving a large number of agents or test parameters. This work intends to apply parallel computing concepts to treat this problem. To this end, it was developed a system called PS2 E2 , an evolution of a previous work, whose utilization in some scenarios allowed the verification of the influence of some parameters such as the population size and the expressiveness of the strategy representation model in the global utility of a society of agents that play the n-Player Prisoner Dilemma.

O Vírus da Imunodeficiência Humana (HIV) é o agente etiológico da Síndrome da Imunodeficiência Adquirida (AIDS). A AIDS é uma doença de distribuição mundial, e estima-se que existam atualmente pelo menos 36,9 milhões de pessoas infectadas com o vírus. Durante o seu ciclo replicativo, o HIV promove diversas alterações na fisiologia da célula hospedeira a fim de promover sua sobrevivência e potencializar a replicação. A rápida progressão da infecção pelo HIV-1 em humanos e em modelos animais está intimamente ligada à função da proteína acessória Nef. Dentre as diversas ações de Nef está a regulação negativa de proteínas importantes na resposta imunológica, como o receptor CD4. Sabe-se que esta ação resulta da indução da degradação de CD4 em lisossomos, mas os mecanismos moleculares envolvidos ainda são totalmente elucidados. Nef forma um complexo tripartite com a cauda citosólica de CD4 e a proteína adaptadora 2 (AP-2), em vesículas revestidas por clatrina nascentes, induzindo a internalização e degradação lisossomal de CD4. Pesquisas anteriores demonstraram que o direcionamento de CD4 aos lisossomos por Nef envolve a entrada do receptor na via dos corpos multivesiculares (MVBs), por um mecanismo atípico, pois, embora não necessite da ubiquitinação de carga, depende da ação de proteínas que compõem os ESCRTs (Endosomal Sorting Complexes Required for Transport) e da ação de Alix, uma proteína acessória da maquinaria ESCRT. Já foi reportado que Nef interage com subunidades dos complexos AP-1, AP-2, AP-3 e Nef não parece interagir com subunidades de AP-4 e AP-5. Entretanto, o papel da interação de Nef com AP-1 e AP-3 na regulação negativa de CD4 ainda não está totalmente elucidado. Ademais, AP-1, AP-2 e AP-3 são potencialmente heterogêneos devido à existência de isoformas múltiplas das subunidades codificadas por diferentes genes. Todavia, existem poucos estudos para demonstrar se as diferentes combinações de isoformas dos APs são formadas e se possuem propriedades funcionais distintas. O presente trabalho procurou identificar e caracterizar fatores celulares envolvidos na regulação do tráfego intracelular de proteínas no processo de regulação negativa de CD4 induzido por Nef. Mais especificamente, este estudo buscou caracterizar a participação do complexo AP-1 na modulação negativa de CD4 por Nef de HIV-1, através do estudo funcional das duas isoformas de ?-adaptina, subunidades de AP-1. Utilizando a técnica de Pull-down demonstramos que Nef é capaz de interagir com ?2. Além disso, nossos dados de Imunoblot indicaram que a proteína ?2-adaptina, e não ?1-adaptina, é necessária no processo de degradação lisossomal de CD4 por Nef e que esta participação é conservada para degradação de CD4 por Nef de diferentes cepas virais. Ademais, por citometria de fluxo, o silenciamento de ?2, e não de ?1, compromete a diminuição dos níveis de CD4 por Nef da membrana plasmática. A análise por imunofluorêsncia indireta também revelou que a diminuição dos níveis de ?2 impede a redistribuição de CD4 por Nef para regiões perinucleares, acarretando no acúmulo de CD4, retirados por Nef da membrana plasmática, em endossomos primários. A depleção de ?1A, outra subunidade de AP-1, acarretou na diminuição dos níveis celulares de ?2 e ?1, bem como, no comprometimento da eficiente degradação de CD4 por Nef. Além disso, foi possível observar que, ao perturbar a maquinaria ESCRT via super-expressão de HRS (uma subunidade do complexo ESCRT-0), ocorreu um acumulo de ?2 em endossomos dilatados contendo HRS-GFP, nos quais também detectou-se CD4 que foi internalizado por Nef. Em conjunto, os resultados indicam que ?2-adaptina é uma importante molécula para o direcionamento de CD4 por Nef para a via ESCRT/MVB, mostrando ser uma proteína relevante no sistema endo-lisossomal. Ademais, os resultados indicaram que as isoformas ?-adaptinas não só possuem funções distintas, mas também parecem compor complexos AP-1 com diferentes funções celulares, já que apenas a variante AP-1 contendo ?2, mas não ?1, participa da regulação negativa de CD4 por Nef. Estes estudos contribuem para o melhor entendimento dos mecanismos moleculares envolvidos na atividade de Nef, que poderão também ajudar na melhor compreensão da patogênese do HIV e da síndrome relacionada. Em adição, este trabalho contribui para o entendimento de processos fundamentais da regulação do tráfego de proteínas transmembrana no sistema endo-lisossomal.
The Human Immunodeficiency Virus (HIV) is the etiologic agent of Acquired Immunodeficiency Syndrome (AIDS). AIDS is a disease which has a global distribution, and it is estimated that there are currently at least 36.9 million people infected with the virus. During the replication cycle, HIV promotes several changes in the physiology of the host cell to promote their survival and enhance replication. The fast progression of HIV-1 in humans and animal models is closely linked to the function of an accessory protein Nef. Among several actions of Nef, one is the most important is the down-regulation of proteins from the immune response, such as the CD4 receptor. It is known that this action causes CD4 degradation in lysosome, but the molecular mechanisms are still incompletely understood. Nef forms a tripartite complex with the cytosolic tail of the CD4 and adapter protein 2 (AP-2) in clathrin-coated vesicles, inducing CD4 internalization and lysosome degradation. Previous research has demonstrated that CD4 target to lysosomes by Nef involves targeting of this receptor to multivesicular bodies (MVBs) pathway by an atypical mechanism because, although not need charging ubiquitination, depends on the proteins from ESCRTs (Endosomal Sorting Complexes Required for Transport) machinery and the action of Alix, an accessory protein ESCRT machinery. It has been reported that Nef interacts with subunits of AP- 1, AP-2, AP-3 complexes and Nef does not appear to interact with AP-4 and AP-5 subunits. However, the role of Nef interaction with AP-1 or AP-3 in CD4 down-regulation is poorly understood. Furthermore, AP-1, AP-2 and AP-3 are potentially heterogeneous due to the existence of multiple subunits isoforms encoded by different genes. However, there are few studies to demonstrate if the different combinations of APs isoforms are form and if they have distinct functional properties. This study aim to identify and characterize cellular factors involved on CD4 down-modulation induced by Nef from HIV-1. More specifically, this study aimed to characterize the involvement of AP-1 complex in the down-regulation of CD4 by Nef HIV-1 through the functional study of the two isoforms of ?-adaptins, AP-1 subunits. By pull-down technique, we showed that Nef is able to interact with ?2. In addition, our data from immunoblots indicated that ?2- adaptin, not ?1-adaptin, is required in Nef-mediated targeting of CD4 to lysosomes and the ?2 participation in this process is conserved by Nef from different viral strains. Furthermore, by flow cytometry assay, ?2 depletion, but not ?1 depletion, compromises the reduction of surface CD4 levels induced by Nef. Immunofluorescence microscopy analysis also revealed that ?2 depletion impairs the redistribution of CD4 by Nef to juxtanuclear region, resulting in CD4 accumulation in primary endosomes. Knockdown of ?1A, another subunit of AP-1, resulted in decreased cellular levels of ?1 and ?2 and, compromising the efficient CD4 degradation by Nef. Moreover, upon artificially stabilizing ESCRT-I in early endosomes, via overexpression of HRS, internalized CD4 accumulates in enlarged HRS-GFP positive endosomes, where co-localize with ?2. Together, the results indicate that ?2-adaptin is a molecule that is essential for CD4 targeting by Nef to ESCRT/MVB pathway, being an important protein in the endo-lysosomal system. Furthermore, the results indicate that ?-adaptins isoforms not only have different functions, but also seem to compose AP-1 complex with distinct cell functions, and only the AP-1 variant comprising ?2, but not ?1, acts in the CD4 down-regulation induced by Nef. These studies contribute to a better understanding on the molecular mechanisms involved in Nef activities, which may also help to improve the understanding of the HIV pathogenesis and the related syndrome. In addition, this work contributes with the understanding of primordial process regulation on intracellular trafficking of transmembrane proteins.

Trabalho de projecto de mestrado em Engenharia Informática, apresentado à Universidade de Lisboa, através da Faculdade de Ciências, 2008
Está a ser desenvolvido, na unidade de investigação LabMAg, o projecto “AutoFocus:Adaptive Self-Improving Multi-Agent Systems", no qual o presente trabalho de mestrado se enquadra. O projecto AutoFocus tem como objectivo a implementação de sistemas multi-agente baseados em entidades autonómicas capazes de comportamentos auto-optimizados e adaptativos. A noção de computação autonómica, tal como outras noções que também implicam computação pró-activa, baseia-se em entidades autónomas que agem activamente no sentido de alcançar os seus objectivos e que têm a capacidade de se adaptar dinamicamente a mudanças no seu ambiente, restringidas por limites de tempo e de recursos. Na abordagem do projecto AutoFocus essa adaptação à mudança, assim como a regulação das capacidades dos agentes, é resultante da combinação de aspectos cognitivos com aspectos de base emocional. O modelo de agente subjacente ao projecto AutoFocus é o Modelo de Agente de Fluxo. A tarefa a que correspondeu este projecto de mestrado, consistiu em desenvolver uma plataforma de implementação para o Modelo de Agente de Fluxo. Pretendeuse com esta plataforma disponibilizar uma ferramenta que permita a rápida implementação de agentes baseados neste modelo bem como a sua monitorização. O trabalho desenvolvido consistiu na análise e desenho, orientado a objectos, implementação e teste dos componentes desta plataforma.
The work presented in this document is part of the project “AutoFocus: Adaptive Self-Improving Multi-Agent Systems" that is being developed at the research unit LabMAg, which objective is the implementation of multi-agent systems based on autonomous entities capable of self-optimized and adaptive behaviors. The notion of autonomic computation, like other notions that also imply proactive computation, is based on autonomous entities that actively work to achieve their objectives and have the ability to dynamically adjust to changes in their environment, constrained by time and resource limits. In the approach used by the AutoFocus project, that adaptation to change and the regulation of the agent's capabilities, result from the combination of cognitive aspects with emotional based aspects. The agent model de_ned and used by the AutoFocus project is the Agent Flow Model. The task that corresponded to the work presented in this document was to develop a platform for the Agent Flow Model. It was intended, with this platform, to provide a tool that enables the rapid deployment and monitoring of agents based on this model. The developed work consisted in the analysis and design, oriented to objects, implementation and testing of components of this platform.

碩士
國立雲林科技大學
資訊管理研究所
88
The quality and quantity of the supply and demand of information are getting more important in recent years. Previous researches have focused on searching, retrieving, and mining useful information for years. However, as the demand of information is getting complicated, traditional tools to accomplish these objectives are facing problems of adaptability. This research proposes an artificial world model to explore the phenomenon of the supply and demand of information in the real world. An adaptive information agent (AIA) is also shown to observe the interaction between it and its environment. Our experiment results show that group selection strategy of information individuals tends to be superiority-sensitive if the information weights of strong and weak information are clearly identified. It is also shown that the learning time tends to be inverted U-shaped as the increasing similarity of environment is easily identified, otherwise, the learning time decreases as the probability of the environment similarity increases. Besides, strong information contributes more than weak information does to the learning of the agent, while weak information, in addition, plays an important role in the learning process of the agent. Though the learning speed is lower down as the weights of strong and weak information become close, the adaptability of each agent is getting clearly seen instead.

M.Sc. (Information Technology)
The purpose of the research presented in this dissertation is to explore the uses of an adaptive multi-agent system for critical information infrastructure protection (CIIP). As the name suggests, CIIP is the process of protecting the information system which are connected to the infrastructure essential to the continued running of a country or organisation. CIIP is challenging due largely to the diversity of these infrastructures. The dissertation examines a number of artificial intelligence techniques that can be applied to CIIP; these techniques range from multi-agent systems to swarm optimisation. The task of protection is broken into three distinct areas: preventing unauthorised communication from outside the system; identifying anomalous actions on computers within the system; and ensuring that communication within the system is not modified externally. A multi-agent learning model, MALAMANTEAU, is proposed as a way to address the problem of CIIP. Due to various problems facing CIIP, multi-agent systems present good opportunities for solving these many problems in a single model. Agents within the MALAMANTEAU model will use diverse artificial and computational intelligence techniques in order to provide an adaptable approach to protecting critical networks. The research presented in the dissertation shows how computational intelligence can be employed alongside multi-agent systems in order to provide powerful protection for critical networks without exposing further security risks.

In the domain of dynamically evolving distributed systems composed of autonomous and (self-) adaptive components, the task of systematically managing the design complexity of their communication and composition is a pressing issue. This stems from the dynamic nature of such systems, where components and their bindings may appear and disappear without anticipation. As one way of addressing the challenge, [15] proposes a new component model (called DEECo), which features separation of concerns via a mechanism of dynamic implicit bindings with implicit communication. This leads to introduction of autonomous components and their dynamic interaction groups (called ensembles). The goal of the thesis is to reify the DEECo concepts and paradigms in a Java-based implementation connected to a simulation of a virtual world. As such, the thesis should provide a platform for experimentation with DEECo-based applications.

abstract: In order to deploy autonomous multi-robot teams for humanitarian demining in Colombia, two key problems need to be addressed. First, a robotic controller with limited power that can completely cover a dynamic search area is needed. Second, the Colombian National Army (COLAR) needs to increase its science, technology and innovation (STI) capacity to help develop, build and maintain such robots. Using Thangavelautham's (2012, 2017) Artificial Neural Tissue (ANT) control algorithm, a robotic controller for an autonomous multi-robot team was developed. Trained by a simple genetic algorithm, ANT is an artificial neural network (ANN) controller with a sparse, coarse coding network architecture and adaptive activation functions. Starting from the exterior of open, basic geometric grid areas, computer simulations of an ANT multi-robot team with limited time steps, no central controller and limited a priori information, covered some areas completely in linear time, and other areas near completely in quasi-linear time, comparable to the theoretical cover time bounds of grid-based, ant pheromone, area coverage algorithms. To mitigate catastrophic forgetting, a new learning method for ANT, Lifelong Adaptive Neuronal Learning (LANL) was developed, where neural network weight parameters for a specific coverage task were frozen, and only the activation function and output behavior parameters were re-trained for a new coverage task. The performance of the LANL controllers were comparable to training all parameters ab initio, for a new ANT controller for the new coverage task. To increase COLAR's STI capacity, a proposal for a new STI officer corps, Project ÉLITE (Equipo de Líderes en Investigación y Tecnología del Ejército) was developed, where officers enroll in a research intensive, master of science program in applied mathematics or physics in Colombia, and conduct research in the US during their final year. ÉLITE is inspired by the Israel Defense Forces Talpiot program.
Dissertation/Thesis
Doctoral Dissertation Applied Mathematics for the Life and Social Sciences 2019

碩士
國立高雄應用科技大學
資訊管理系碩士在職專班
97
With the rapid development of e-commerce and the use of agents, auto transaction and auto-negotiation between buyer and seller becomes one of the issues to resolve. Continuous Double Auction in the trading mechanism, because of the randomness of the transaction process and transaction pricing face complex decision-making, therefore, receive more attention. In this paper, continuous review of the relevant double auction bidding strategies, we found that adaptive strategies (Adaptive attitude strategy, AA strategy), Ma, Ho (2007), is better than the other bidding strategies. Therefore, AA bidding strategy of traders is selected as main strategy of double auction trading environment. If we can accurately forecast prices, trader will be greater profits. Therefore the use of neural network models and multiple regression model, inference to predict, identify more effective predictive model. The results showed that: 1. Both neural network models or multiple regression model, the more predictable variables, the forecast would have been better performance. 2. Neural network model predictive capability is better than multiple regression model. By the implementation system, this paper proposed the use of trading market as an agent-based of the auction system, to both buyers and sellers, this system can reduce the bid cost of decision-making, but also very fair to carry on their activities. By the experimental results, the neural network forecast model in dealing with non-linear, so, this paper proposed an agent-based auction system using the neural network model to forecast prices.