Multi Agent Systems and Technologies

Actors in competitive environments are bound to decide and act under conditions of uncertainty because they rarely have accurate foreknowledge of how their opponents will respond and when they will respond. Just as a competitor makes a move to improve their standing on a given variable relative to a target competitor, she should expect the latter to counteract with an iterative lagged asymmetric response, that is, with a sequence of countermoves (iteration) that is very different in kind from its trigger (asymmetry) and that will be launched at some unknown point in the future (time lag). The paper explicates the broad relevance of the newly proposed concept of " iterative lagged asymmetric responses " to the social study of temporality and to fields as diverse as intelligence and counterintelligence studies, strategic management, futures studies, military theory, and long-range planning. By bringing out in the foreground and substantiating the observation that competitive environments place a strategic premium on surprise, the concept of iterative lagged asymmetric responses makes a contribution to the never-ending and many-pronged debate about the extent to which the future can be predicted.
DOI: https://doi.org/10.1177/0961463X17752652

Whether collective agency is a coherent concept depends on the theory of agency that we choose to adopt. We argue that the enactive theory of agency developed by Barandiaran, Di Paolo and Rohde (2009) provides a principled way of grounding agency in biological organisms. However the importance of biological embodiment for the enactive approach might lead one to be skeptical as to whether artificial systems or collectives of individuals could instantiate genuine agency. To explore this issue we contrast the concept of collective agency with multi-agent systems and multi-system agents, and argue that genuinely collective agents instantiate agency at both the collective level and at the level of the component parts. Developing the enactive model, we propose understanding agency - both at the level of the individual and of the collective - as spectra that are constituted by dimensions that vary across time. Finally, we consider whether collectives that are not merely metaphorically 'agents' but rather are genuinely agentive also instantiate subjectivity at the collective level. We propose that investigations using the perceptual crossing paradigm suggest that a shared lived perspective can indeed emerge but this should not be conflated with a collective first person perspective, for which material integration in a living body may be required.

In this paper, we have described a solution for supporting independent living of the elderly by means of equipping their home with a simple sensor network to monitor their behaviour. Standard home automation sensors including movement sensors and door entry point sensors are used. By monitoring the sensor data, important information regarding any anomalous behaviour will be identified. Different ways of visualizing large sensor data sets and representing them in a format suitable for clustering the abnormalities are also investigated. In the latter part of this paper, recurrent neural networks are used to predict the future values of the activities for each sensor. The predicted values are used to inform the caregiver in case anomalous behaviour is predicted in the near future. Data collection, classification and prediction are investigated in real home environments with elderly occupants suffering from dementia.

The proper system for evaluating the learning achievement of students is the key to realizing the purpose
of education. In recent years, several methods have been presented for applying the fuzzy set theory in
the educational grading systems. In this paper, we propose a method for the evaluation of students’
answerscripts using a fuzzy system. The proposed system applies fuzzification, fuzzy inference, and
defuzzification in considering the difficulty, the importance and the complexity of questions. The transparency,
objectivity, and easy implementation of the proposed fuzzy system provide a useful way to
automatically evaluate students’ achievement in a more reasonable and fairer manner.
 2008 Elsevier Ltd. All rights reserved.

In this paper, we present an application of artificial neural networks to the real-world problem of predicting forest fires. The neural network used for this application is a multilayer perceptron whose architectural parameters, i.e., the number of hidden layers and the number of neurons per layer were heuristically determined. The synaptic weights of this architecture were adjusted using the backpropagation learning algorithm and a large set of real data related to the studied problem. We also present and discuss some examples of illustrating results that show the performance and the usefulness of the resulting neural system.

This paper presents an innovative soft computing architecture based on a combination of DAI (distributed artificial intelligence), fuzzy-genetic driven embedded-agents and IP Internet technology applied to the domain of intelligent-buildings. It describes the nature of intelligent buildings (IB) and embedded-agents, explaining the unique control and learning problems they present. We show how fuzzy -logic techniques can be used to create a behaviour-based multi-agent architecture in intelligent-buildings. We discuss how this approach deals with the highly unpredictable and imprecise nature of the physical world in which the system is situated, and how embedded-agents can be constructed that utilise sensory information to learn to perform tasks related to user comfort, energy conservation, and safety. We explain in detail our machine learning methodology that is based on a novel genetic algorithm mechanism referred to as an associative experience engine (AEE) and present the results of practical experiments. We compare results obtained from the AEE approach to that of the widely known Mendel-Wang method. Finally we explain potential applications for such systems ranging from commercial buildings to living-area control systems for space vehicles and planetary habitation modules.

In this paper, we describe a novel system for learning and adapting type-2 fuzzy controllers for intelligent agents that are embedded in Ubiquitous Computing Environments (UCE). Our type-2 agents learn online in a non intrusive manner and in a life long learning mode the user behaviours to control the UCE on the user’s behalf. We have performed unique experiments in which the type-2 intelligent agent has learnt and adapted online to the user’s behaviour during a stay of five days in the intelligent Dormitory (iDorm) which is a real ubiquitous computing environment test bed. We will show how our type-2 agents deal with the uncertainty and imprecision present in UCE to give a very good response that outperforms the type-1 fuzzy agents while using a smaller number of rules.

In this paper, we describe a new application domain for intelligent autonomous systems – Intelligent Buildings (IB). In doing so we present a novel approach to the implementation of IB based on a hierarchical fuzzy genetic multi embedded-agent architecture comprising a low-level behaviour based reactive layer whose outputs are coordinated in a fuzzy way according to deliberative plans. The fuzzy rules related to resident’s comfort are learnt online in a short time interval using a novel Fuzzy-Genetic technique from the resident’s actual behaviour in a learning phase. Our approach utilises an intelligent agent approach to autonomously governing the building environment. We discuss the role of learning in building control systems, and contrast this approach with existing IB solutions. We explain the importance of acquiring information from sensors, rather than relying on preprogrammed models, to determine user needs. We describe how our architecture, consisting of distributed embedded agents, utilises sensory information to learn to perform tasks related to user comfort, energy conservation, and safety. We show how these agents, employing a behaviour-based approach derived from robotics research ("a building is a robot we live inside"), are able to continuously learn and adapt to individuals within a building, whilst always providing a fast, safe response to any situation. Such a system could be used to provide support for older people, or people with disabilities, allowing them greater independence and quality of life.

In this paper the problem of cyclic instability in dynamic environments is presented. This cyclic instability is generated when binary rule-based nomadic agents (agents entering or leaving the environment) interact in complex ways, generating undesirable outputs for the final user. Our strategy is focused on minimizing this cyclic behaviour, using optimization algorithms, in particular Genetic and Differential Evolution Algorithms. These algorithms are applied to the Average Change Function. Different test instances were used to evaluate the performance of these algorithms. Additionally, statistical tests were applied to measure their performance.

An amorphous computer is a multitude of tiny computers each with a CPU, memory, and local communication capability. An iSurface is a particular instance of an amorphous computer, an “intelligent” coasting capable of computing, sensing and some limited output. This can be regarded as a massively parallel computing system, albeit one with random structure and unreliable nodes, and, as such, is likely to be extremely difficult to program conventionally. Prior research into amorphous computing suggests that utilising principles of self-organisation is successful in providing some functionality to an amorphous computer. Here we argue there is a definite need for techniques in engineering emergent functionality to match user requirements; the development of systems that utilise self-organisation in a user-directed manner. We present an early version of a methodology to do this and an example of its successful usage.

Inhabited environments offer a diverse set of problem domains which can benefit from the application of artificial intelligence (AI). This paper presents a work in progress conducted between British Telecom (BT) and the University of Essex. The work argues that a single knowledge representation and processing model (followed by traditional AI systems) is not adequate to fulfill the requirements of all inhabited environment problem domains (especially with near real-time temporal complexity). Based on this predicate, the work seeks to explore the use of multiple knowledge languages which each address a specific problem domain.

The interdependence and complexity of socio-technical systems and availability of a wide variety of policy measures to address policy problems make the process of policy formulation difficult. In order to formulate sustainable and efficient transport policies, development of new tools and techniques is necessary. One of the approaches gaining ground is policy packaging, which shifts focus from implementation of individual policy measures to implementation of combinations of measures with the aim of increasing efficiency and effectiveness of policy interventions by increasing synergies and reducing potential contradictions among policy measures. In this paper, we describe the development of a virtual environment for the exploration and analysis of different configurations of policy measures in order to build policy packages. By developing systematic approaches it is possible to examine more alternatives at a greater depth, decrease the time required for the overall analysis, provide real-time assessment and feedback on the effect of changes in the configurations, and ultimately form more effective policies. The results from this research demonstrate the usefulness of computational approaches in addressing the complexity inherent in the formulation of policy packages. This new approach has been applied to the formulation of policies to advance sustainable transportation.

Contracts are designed to govern the relations between business partners and allocate risk among them, yet they cannot mitigate all risks; hence, dispute resolution mechanisms have been developed to assist. According to research, arbitration is considered one of the most efficient legally binding dispute resolution mechanisms. Since initiating an arbitration case against business partners often severs the business relationship between them, predicting the arbitrator's decision becomes valuable to the arbitrating parties. To do so, arbitration prediction models are suggested in this paper through modeling the arbitrator's award in two stages: evidence admissibility and evidence evaluation. While arbitrators currently admit evidence in a relatively unsystematic way, this proposed admissibility framework systemizes the current practice by preserving its subjectivity, yet aligning it with the widely deployed rules of arbitration. This research further proposes an evidence evaluation model, providing a rapid answer to the famous commercial arbitration battle, discovery. While the current arbitrators' practice in allowing discovery is not systematic, the proposed optimization model is a tool that could systematically assist arbitrators trading between the arbitration key characteristic, efficiency, and justice. This is achieved by assigning weights to evidence parameters such as sufficiency and credibility, which in turn provide numerical measures to reduce the decision's subjectivity.

Information Systems (IS) play a central part in all aspects of our world, from science, medicine, and engineering, to industry, business, law, politics, government, arts, culture, society, health, operational support in daily life, surveillance of infrastructure and environment, and homeland protection and national security. Without proper security precautions, IS are prone to unexpected and intolerable side effects, such as leakage of operational and confidential data, identity theft and unauthorized access, and possibly modification of private data, services, and systems. Security services are required in order to guarantee information security and protection of privacy, such as data confidentiality, data authentication, anonymity, entity authentication, non-repudiation of origin and receipt, access control, protection against denial of service and phishing, and secure processing and deletion of data. In a nutshell, dependable and trustworthy security solutions based on strong cryptography are needed.

Multi-Agent systems (MAS) consist of multiple intelligent agents that interact to solve problems that may be beyond the capabilities of a single agent or system. For many years, conceptual MAS designs and architectures have been proposed for applications in power systems and power engineering.  With the increasing use and modeling of distributed energy resources for microgrid applications, MAS are well suited to manage the size and complexity of these energy systems. The purpose of this paper is to survey applications of MAS in the control and operation of microgrids.  The paper will review MAS concepts, architectures, develop platforms and processes, provide example applications, and discuss limitations.

In this paper we introduce the concept of an iSurface; a surface coated with a multitude of identical nano-scale computing devices called iCells. Based on earlier extensive simulation work, we describe the computational limits of amorphous computing for image display applications and explain how such surfaces could be used to create a variety of novel fashion applications ranging from electronic wallpaper, active jewellery to adaptive clothing.

In this paper we describe a novel multi-agent architecture consisting of two remotely (geographically dispersed) associated intelligent environments aimed at establishing an agent based care system. We will introduce the 'Care Agents' joint project, in which a collection of heterogeneous building and robotic embedded agents intelligently learns and cooperates to care for human occupants. We will show how our agent architecture facilitates inter-agent communication and how this system can be used to detect abnormalities from normal.

Wireless Sensor Networks (WSN) are a fast-emerging area of interest in the domain of large-scale communication networks. However, design of novel WSN applications requires first developing models and performing extensive
simulation. Modern large-scale WSNs are expected to be both dynamic as well as random and often spread over a large scale. However, this problem has not previously been addressed in literature. In our previous works, we have demonstrated how Agent-based modeling may be used to effectively model various types of Complex Adaptive Systems such as, but not limited to, Self-organizing Communication infrastructures. In this paper, we present first steps towards providing a comprehensive set
of guidelines and tutorial for developing deployment models of WSNs besides modeling routing algorithms using agent-based modeling. Our simulation results demonstrate the effectiveness and ease of use coupled with a short learning curve involved in developing agent-based models of complex WSN applications.

We outline an approach based on using embedded agents for realising ambient environmental intelligence within a ubiquitous computing environment. We describe a practical ubiquitous computing test-bed environment, the Essex intelligent Dormitory (iDorm), which combines a number of heterogeneous computational artefacts and networks. We show novel results where a user occupied the iDorm for five and half days. During this period the iDorm was monitored and controlled by an intelligent and adaptive embedded agent using our fuzzy logic based Incremental Synchronous Learning (ISL) approach. The agent operates in a life-long learning mode realising the ambient intelligence vision in which it learns the user’s behaviour and adapts to his needs and controls the iDorm according to the user’s preferences in a personalised and non-intrusive way.

Intisari Pesatnya perkembangan pasar global menimbulkan persaingan yang begitu nyata pada semua sisi perusahaan, diperlukan perhitungan dan perencanaan yang benarbenar matang untuk dapat ikut bersaing atau bahkan memenangkan persaingan pasar global tersebut. Hal ini membuat pertumbuhan perusahaan-perusahaan penyedia jasa manajemen bisnis dan pembuat aplikasi manajemen perusahaan semakin berkembang dengan pesat juga, hingga menghasilkan metode-metode terbaru yang bisa diandalkan. Salah satunya adalah perusahaan SAP yang sudah mendunia dibidang ini. Dan salah satu bidang cakupan yang di tangani oleh SAP adalah perencanaan sumber daya perusahaan dengan aplikasi dan metode andalannya yang sering disebut dengan Enterprise Resource Planning (ERP-SAP).

The dynamic nature of grid resources and the demands of users produce complexity in the grid scheduling problem that cannot be addressed by deterministic algorithms with polynomial complexity. One of the best methods for grid scheduling is the genetic algorithm (GA); the simple and parallel features of this algorithm make it applicable to several optimization problems. A GA searches the problem space globally and is unable to search locally. Therefore, scholars have investigated combining GAs with other meta-heuristic methods to resolve the local search problem.  This is the focus of the present contribution, where we have developed a new hybrid scheduling algorithm that combines a GA and the gravitational emulation local search (GELS) algorithm denotes GGA. The noteworthy feature of the proposed optimal scheduler is that it decreases run-time and the number of submitted tasks whose deadlines are missed. A comparison of the performance of our proposed joint optimal scheduler to similar methods shows that it produces more optimal computation time.

several neural networks controllers for robotics manipulators have been developed during the last decades due to their capability to learn the dynamics properties and the improvements in the global stability of the system. In this paper, two control and identification schemes for a two links robotic manipulator implementing neural networks are presented. A multilayer feedforward neural network and a model reference adaptive controller are used to estimate the inverse dynamic of this mechanism.
An online training algorithm based on the error dynamics is used on the adaptive neural network controller but the neural networks are trained offline with a backpropagation algorithm. The design and architecture of the neural networks are explained along with the identification procedure of the robotic system. Simulations and comparisons with a PD controller are done to test the performance of the neural network controller.

Increasing performance and decreasing cost of microprocessors are making it feasible to move more processing power to the data source. This allows us to investigate new methods of storage delivery and storage management that were not plausible in the past. Our architecture, inspired by agent-based techniques and active disk technology, promotes an open storage management framework that embeds functionality into storage devices. We use local agents to implement self-control with automated capability that can be dynamically adapted to meet the storage management through improved search capabilities in the virtual environment and also retain capabilities like: security, performance and availability requirements.

Con questa Tesi di Laurea si muovono i primi passi di una ricerca applicata finalizzata alla costruzione-deposizione di materiale da parte di sciami di mini-robot dal comportamento indipendente che si coordinano tramite segnali lasciati e rilevati nell'ambiente in cui si muovono.

Lo sviluppo di tecniche di progettazione e fabbricazione digitale ha prodotto un aumento nel grado di interconnessione tra tecnologia e design, dunque, di nuove possibilità tettoniche. Le relazioni tettoniche tradizionali stanno infatti subendo una trasformazione radicale, potendo essere esplicitamente informate e dunque mediate attraverso gli strumenti digitali dall’ideazione alla produzione. Questa mediazione informata del contenuto tettonico (che opera costantemente) è distintivo di un approccio material-based alla progettazione che aumenta l’integrazione tra struttura, materia e forma entro le tecnologie di fabbricazione (R.Oxman).

Dei numerosi processi di fabbricazione per l’architettura che si servono di tecnologia robotica, pochi sono capaci di superare la logica gerarchica, rigida e lineare-sequenziale che serve di fatto agli obiettivi di automazione ed ottimizzazione. La distribuzione di forme di intelligenza semplificata ad un numero elevato di unità robot è quindi qui proposta come alternativa al modello appena descritto. Incorporando semplici decisioni di carattere architettonico negli agenti-robot che costituiscono il sistema distribuito di entità autonome, la loro interazione e le decisioni prese individualmente producono comportamento collettivo e l’integrazione delle suddette relazioni tettoniche.

Nello sviluppo del progetto, si è fatto così riferimento a modelli comportamentali collettivi (di sciame) osservabili in specie comunitarie che organizzano strutture materiali -come termiti e vespe- ed in organismi semplici -come le muffe cellulari della specie Physarum polycephalum.
Per queste specie biologiche il processo di costruzione non dipende da un 'piano generale' ma è guidato esclusivamente da azioni dei singoli individui che comunicano lasciando tracce chimiche nell'ambiente e modificano il loro comportamento rilevando le tracce lasciate dagli altri individui.

A questo scopo, oltre alle simulazioni in digitale, è stato indispensabile sviluppare dei prototipi funzionali di tipo fisico, ovvero la realizzazione di mini-robot dal movimento indipendente, in grado di coordinarsi tra loro tramite segnali lasciati nell'ambiente e capaci di depositare materiale.

Lo sviluppo della costruzione creata dallo sciame di robot prevede un processo di crescita verticale dal basso verso l'alto conseguente a cicli di rilascio del materiale trasportato a bordo delle unità robot ed indurimento dello stesso sul piano di deposito.
La simulazione ha cercato di considerare per quanto possibile i numerosi vincoli derivanti dalla realizzazione del prototipo funzionale, e viceversa, la progettazione delle componenti hardware si è ispirata ai modelli della simulazione.

Pur considerato lo stato embrionale del prototipo di tecnologia proposto, sono numerosi i feedback forniti dall’esperienza di ricerca: a partire da questi, un’ulteriore fase di approfondimento del progetto saprà intervenire in modo mirato alla correzione dell’architettura stessa del processo, ed al miglioramento della robotica impiegata fin qui a dimostrazione di fattibilità.

This is a user’s guide for the Spice-SOM - a Self-Organizing Map (SOM) application. It doesnot intend to introduce about SOM theories. You can find more knowledge about SOM and Neural Network (NN) in other textbooks.Depending on the version, some contents of this material may be different with your downloaded Spice-SOM.The purpose of this program is to get you started quickly with Neural Network without havingto go through lengthy theory of the Neural Network background. Once you understand these programs you will be able to consult the Neural Network materials on a need basis.Spice-SOM's arm is to introduce NN and SOM to graduated students studying ComputationalIntelligence. Currently Spice-SOM has been using by many students around the world. Spice-SOM has interfaces in Vietnamese, English and Japanese.Spice-SOM was written by CAO THANG when he did researches in the Soft IntelligenceLaboratory, Ritsumeikan University, Japan, 2003-2007.Spice-SOM and Spice-Neuro can be downloaded at download.cnet.com

As computing devices become more pervasive, our daily activities start generating a vast amount of information that could be exploited for helping us better understand ourselves. In this paper we present a system that uses easily available data correlated into a story-based representation aimed, at providing users with a better understanding of their lifestyles. While this work is still in its early phases, we believe that it provides valuable insights into the design of such systems. Our initial findings show that user data generated through a person’s daily activities can reveal a wealth of valuable information which they can use to adjust and improve their lifestyles.

From the multi-agent system perspective Intelligent Buildings (IB) are a new research area and this domain poses several interesting challenges. Our IB model comprises a building containing room based embedded-agents which control environmental variables and devices within rooms that communicate with each other via a network. Inter-agent communication is central to such multi-agent systems. In this paper we explain why existing agent communication languages (ACLs) are unsuitable for IB embedded-agents and present a specification for a hierarchical Distributed Intelligent Building Agent Language (DIBAL) that overcomes the problems in applying ACLs to IB based embedded-agents. The paper begins by reviewing existing Agent Communication Languages (ACL) and discussing their unsuitability for IB applications before presenting our IB embedded-agent communication language, DIBAL. We then illustrate the ways in which DIBAL would facilitate the functionality we require by looking at a few IB scenarios in some detail.

Inspired by the new achievements in mobile robotics having as a result mobile robots able to execute different production tasks, we consider a factory producing a set of distinct products via or with the additional help of mobile robots. This particularly flexible layout requires the definition and the solution of a complex planning and scheduling problem. In order to minimize production costs, dynamic determination of the number of robots for each production task and the individual robot allocation are needed. We propose a solution in terms of a two-level decentralized Multi-Agent System (MAS) framework: at the first, production planning level, agents are tasks which compete for robots (resources at this level); at the second, scheduling level, agents are robots which reallocate themselves among different tasks to satisfy the requests coming from the first level. An iterative auction based negotiation protocol is used at the first level while the second level solves a Multi-Robot Task Allocation (MRTA) problem through a distributed version of the Hungarian Method. A comparison of the results with a centralized approach is presented.
Highlights

► Factories producing different goods with additional mobile resources (robots). ► A two-level multi-agent system framework for production planning and scheduling. ► First: allocate robots to fulfill demands with minimum production cost. ► An iterative auction method inspired by Lagrangean relaxation is adopted. ► Second: robots (re)schedule themselves with a distributed version of Hungarian method.

Autonomy of embedded agents in intelligent environments is highly debated topic; while some believe that agents should have very minimal autonomy and should only act as directly instructed by the user, others consider providing agents with autonomy to be an essential aspect to building intelligent environments. This paper reports on the current progress of our project to enable human users and agents to collaborate in managing intelligent environments as a team. We seek to develop an adjustable-autonomy agent in an effort to explore user acceptance of pervasive computing and the use of autonomous agents therein, as wells as aiming to improve the robustness and reliability of future intelligent environment systems. We present our Adjustable-autonomy Behaviour-Based Agent (ABBA) architecture model and discuss our initial trials with our prototype system, built on a smart home emulator, which demonstrate the plausibility of employing adjustable autonomy in full-scale intelligent environments and pervasive computing systems.

Multi-agent systems underpin the vision for ambient intelligence. However, developing multi-agent systems is a complex and challenging process. For example, pervasive computing has been found susceptible to instability, due to unwanted behaviour arising from unplanned interaction between rule based agents. This instability is impossible to predict, as it depends on the rules of interaction, the initial state of the system, the user interaction, and in the time delay of the system (due to network traffic, different speed of processing, etc). In this paper we present a theoretical framework, an Interaction Network (IN), together with a communication locking strategy that we call INPRES (Instability Prevention System) that can be used to identify and eliminate this problem. In addition we describe a Multi-Dimensional Model (MDM) to represent the agents and the state of each agent over time. A theorem showing the role of delays in an unstable system is presented. We present experimental results based on simulations and a physical emulation that demonstrate the effectiveness of these methods.

The conjugate gradient optimization algorithm is combined with the modified back propagation algorithm to yield a computationally efficient algorithm for training multilayer perceptron (MLP) networks (CGFR/AG). The computational efficiency is enhanced by adaptively modifying initial search direction as described in the following steps: (1) Modification on standard back propagation algorithm by introducing a gain variation term in the activation function, (2) Calculation of the gradient descent of error with respect to the weights and gains values and (3) the determination of a new search direction by using information calculated in step (2). The performance of the proposed method is demonstrated by comparing accuracy and computation time with the conjugate gradient algorithm used in MATLAB neural network toolbox. The results show that the computational efficiency of the proposed method was better than the standard conjugate gradient algorithm.

The main thesis of this work is that human interactions are neither unpredictable nor bounded, but they are undertaken autonomously on the grounds of a number of basic principles and conditions. Among these, a crucial role is played by the objective dependence relationships holding among agents. In this paper we report about a first step in providing a computational theory of dependence as a tool for interaction control. We define non social as well as social dependence, and try to show how dependence relationships are organized into complex patterns (such as multiparty, multigoal, unilateral, and bilateral dependence). We then show how a given set of dependence relationships may produce new dependence relationships. Finally, we explore the relationship between dependence and influencing, describing how an agent's dependence on another is predictive of one's goal of influencing the other, as well as of the latter's power of influencing the former.

Noise is a form of a pollutant that is terrorizing the occupational health experts for many decades due to its adverse side-effects on the workers in the industry. Noise-Induced Hearing Loss (NIHL) handicap is one out of many health hazards caused due to excessive exposure to high frequency noise emitted from the machines. A number of studies have been carried-out to find the significant factors involved in causing NIHL in industrial workers using Artificial Neural Networks (ANN). Despite providing useful information on hearing loss, these studies have neglected some important factors. The traditional Back-propagation Neural Network (BPNN) is a supervised Artificial Neural Networks (ANN) algorithm. It is widely used in solving many real time problems in world. But BPNN possesses a problem of slow convergence and network stagnancy. Previously, several modifications were suggested to improve the convergence rate of Gradient Descent Back-propagation algorithm such as careful selection of initial weights and biases, learning rate, momentum, network topology, activation function and 'gain' value in the activation function. This research proposed an algorithm for improving the current working performance of Back-propagation algorithm by adaptively changing the momentum value and at the same time keeping the 'gain' parameter fixed for all nodes in the neural network. The performance of the proposed method known as 'Gradient

UAVTraffic (cro. Višeagentno upravljanje prometom autonomnih bespilotnih letjelica - Multi-agent based Traffic Control of Autonomous Unmanned Aerial Vehicles) was financed by the University of Zagreb from August 2014 until December 2014. The technology of unmanned aerial vehicles has matured for various civil usage scenarios and is already being tested by major industry players including Amazon, Google and DHL. Their usage ranges from agriculture, aerial photography, traffic and ecology to surveillance and package delivery. Having the current trends and predictions about future use of this amazing technology in mind, it seems to be just a matter of time until we will alongside terrestrial (on the ground 2D) and aerial traffic (above 300 meters of altitude) have to control traffic of drones (below 300 meters of altitude). This zone of UAV traffic is currently not well regulated and often forbidden in many countries. The research in UAVTraffic is therefore aimed towards making a first step in developing basic
traffic rules for (autonomous) UAVs by using simulation models and small scale field studies. In accordance with this objective, UAVTraffic will make good use of advanced semantic Web technologies (especially of the ontology of cooperative intelligent transport systems developed at the University of Zagreb), multi-agent system methodology (especially proactive agents, automated planning and negotiation algorithms), technologies for machine vision (especially spatial object recognition), the Internet of Things or more specifically the Internet of Vehicles (a network of various devices like smart phones, watches, goggles, cars but also UAVs and other things that will allow for traffic control).

In this study, motion planning and control scheme for a cooperative transportation system, which consists of a single object and multiple autonomous non-holonomic mobile robots included forklifts, is proposed. Virtual leader-follower formation control strategy is used for the cooperative transportation system. The object is assumed as the virtual leader of the system and the robots carrying the object on their forklifts are considered as follower robots. A smooth path is generated by considering the constraints of the virtual robot. The origin of the coordinate system attached to the center of gravity of the object tracks the generated path. A path for each follower robot is generated to keep the formation structure. The follower robots track their paths. A communication framework is used for the messaging between robots, and asymptotically stable tracking control is used for trajectory tracking. The proposed method is verified with real applications and simulations using Pioneer P3-DX mobile robots and a single object.

Paper videos can be reached from this URL: http://www.youtube.com/playlist?list=PLENSkat0854sw9XIuLI_hz3NzoauVJDaC

Please look at the further study of "Formation-based cooperative transportation by a group of non-holonomic mobile robots"

Expert system Fuzzy rule based systems Business management Knowledge base minimal structure Operations management a b s t r a c t Nowadays firms are required to reach high levels of specialisation in order to increase their competitiveness in complex markets. Knowledge management plays a fundamental role in this process as the correct implementation of strategies is determined by the information transfer and dissemination within the organisation. In this paper, a fuzzy expert system focused on increasing accuracy and quality of the knowledge for decision making is designed. A model based on fuzzy rules to simulate the behavior of the firms, is presented under the assumption of determined input parameters previously detected and an algorithm is developed to achieve the minimal structure of the model. The result is a fuzzy expert system tool, called ESROM, which gives valuable information to help managers to improve the achievement of the strategic objectives of the company. A main contribution of this work it that the system is general and can be adapted to different scenarios.

Named Entity Recognition is a prior task in Natural Language Processing. Named Entity Recognition is a sub task of information extraction and it identifies and classifies proper nouns in to its predefined categories such as person, location, organization, time, date etc. In this document the major focus is given on NER approaches and the work done till now for various languages to identify Named Entities is been discussed. Author have done comparative study to recognize named entity and identified that CRF approach proven best for Indian languages to identify named entity.

Simulating the evolution of cities suggests the ability to understand and model the laws governing urban dynamics. The approach presented in this paper uses a multi-agent system to model the selforganising properties of urban systems. The proposed multi-agent system is a hierarchy of topographic agents (e.g. buildings, roads, rivers or urban blocks) that are built, modified, merged, split, restructured and destroyed over time. Each agent's behaviour is controlled by a set of constraints and associated actions. In the proposed system, only topographic objects are taken into account, assuming a certain independence between society and the urban form (morphology) as well as a delay between the action of the society on the urban form and the feedback of the latter to the former.

There are many arguments for and against the use of autonomous-agents in intelligent environments. Some researchers maintain that it is of utmost importance to give complete control to users, and hence greatly restrict autonomy of agents; whereas, others believe that is it preferable to increase user convenience by allowing agents to operate autonomously on the user’s behalf. While both of these approaches have their distinct merits, they are not suitable for all users. As people’s opinions and concerns regarding agent autonomy are highly individual, depending on a wide range of factors and often changing over time, a much more dynamic approach to agent autonomy is needed. This work explores how it is possible to equip intelligent environments with an adjustable autonomy mechanism, which allows an individual user to increase or decrease agent autonomy in order to find their own comfortable sweet-spot between maintaining/relinquishing control and gaining/losing convenience. This paper presents the Adjustable Autonomy Intelligent Environment (AAIE) model, discusses how adjustable autonomy can be achieved in intelligent environments, and discusses the major findings from a recent online survey and user study, which highlight the major factors and concerns of users that determine their personal preferences towards different levels of autonomy.

In this paper we have projected on the involvement of multi-agent system in medical or health care domain. The objective of this study is to provide future researchers more resourceful and focused review of various research papers in this domain. Multi-agent system is most suitable for healthcare paradigm, as the properties of agent based systems deals with heterogeneous multiple agents. Data distribution and data management in a dynamic and distributed environment with multi-user cooperation, made multi-agent system more significant in this field. The disposition of this paper is classified on the basis of theoretical and application approach. We have tried to cover few relevant papers published on last decade. The main aim of this literature survey is to provide a complete road map on multi agent system based research on medical health care platform.

This material guides you to use Self-Organizing Map (SOM) and MLP Neural Networks (NN) in some practical applications. It was used to introduce NNto some Japanese students. The author thought that it might be useful for the other students so he re-prepared it, for your reference. Because of timelimitation, it may have mistakes. Please tell the author at spiceneuroAT gmail DOT com or http://spiceneuro.wordpress.com

If you are interested in NN and SOM, you can use free software, SpiceSOM and SpiceMLP, downloadhere Some data presented here also already in "Data" folder when you setup SpiceSOM and SpiceMLP software. All results here are modeled by SpiceSOMand SpiceMLP.Thank you

In this paper, we focus on computing a consistent traffic signal configuration at each junction that optimizes multiple performance indices, i.e., multi-objective traffic signal control. The multi-objective function includes minimizing trip waiting time, total trip time, and junction waiting time. Moreover, the multi-objective function includes maximizing flow rate, satisfying green waves for platoons traveling in main roads, avoiding accidents especially in residential areas, and forcing vehicles to move within moderate speed range of minimum fuel consumption. In particular, we formulate our multi-objective traffic signal control as a Multi-Agent System (MAS). Traffic signal controllers have a distributed nature in which each traffic signal agent acts individually and possibly cooperatively in a MAS. In addition, agents act autonomously according to the current traffic situation without any human intervention. Thus, we develop a multi-agent multi-objective Reinforcement Learning (RL) traffic signal control framework that simulates the driver's behavior (acceleration/deceleration) continuously in space and time dimensions. The proposed framework is based on a multi-objective sequential decision making process whose parameters are estimated based on the Bayesian interpretation of probability. Using this interpretation together *Text Table(s) Click here to view linked References

Pervasive intelligent environments, such as digital homes, offer huge potentials for eHealth services. However, there is a question as to how best to manage (i.e. configure and program) these systems. The majority of solutions to-date rely exclusively on either end users or autonomous agents to manage the environment, which may not be suitable for all users and in certain situations, especially in eHealth applications where uses may have very specific requirements that can change considerably over time. In this work-in-progress paper we argue that human-agent teamwork is vital for the adoption of future pervasive computing systems into our everyday lives; we discuss these issues, exposing the problem of exclusive management and we explore a solution, presenting our novel architecture for an adjustable autonomy agent that enables human-agent based teamwork. This work contributes to the PAL project: a collaborative research effort between University of Essex, BT, University of Cambridge and Ericsson.

Quantum analogues of the (classical) Logical Neural Networks (LNN) models are proposed in [6] (q-LNN for short). We shall here further develop and investigate the q- LNN composed of the quantum analogue of the Probabilis- tic Logic Node (PLN) and the multiple-valued PLN (MPLN) variations, dubbed q-PLN and q-MPLN respectively. Be- sides a clearer mathematical description, we present a com- putationally efficient and simply described quantum learn- ing algorithm in contrast to what has been proposed to the quantum weighted version.