Dissertations / Theses on the topic 'Intelligent model'

This thesis focuses on visual servoing (VS) control systems, particularly on image-based visual servoing (IBVS) control structures. In IBVS, the error signal is computed in the image plane and the regulation commands are generated with respect to such error by means of a visual Jacobian. The main design challenge is the high latency of the visual sensor which affects the overall performance and limits the design. The primary objective is to develop a complete framework for simulation and real-time experimentation of VS schemes. One commercial CCD camera is attached to the TQ MA2000 robotic manipulator. The framework has been employed to investigate the use of RL algorithms to increase the performance of the IBVS control structure. The classic RL actor-critic structure has been used to perform on-line adjustment of the gains driving the linear trajectory regulator inside the IBVS control structure. The neural system learns directly from data in the image space and the state of the robot. Two feedforward networks are used, the actor directly modifies the regulator gains whereas the adaptive critic stores and assigns action values. By using the adaptive heuristic critic approach (AHC), the training aims to achieve real-time improvement and adaptation without losing an acceptable regulation of the visual servoing task. A compact model and a flexible framework host the reinforcement learning algorithm in order to enable its inclusion within the IBVS control structure. The approach in this thesis has solved critical neuro-dynamic problems which are derived from the interaction between the imaging model and the robot’s dynamics. The VS toolkit also provides a real-time library to implement and test the IBVS control structure. The libraries have proven effective to construct both the linear IBVS and the RL-supported IBVS system thanks to its layered architecture which facilitates the inclusion of control en› tities of different nature such as the neural networks and the learning framework. Two case studies demonstrate the applicability of the CSC VS toolkit to integrate all the required components and to implement each of the VS experiments in real-time. Performance comparison between the linear IBVS and the RL-supported system are also documented to show the effectiveness of the actor-critic structure.

Submitted by (ana.araujo@ufrpe.br) on 2016-08-03T14:07:18Z No. of bitstreams: 1 Luiz Jose Rodrigues dos Santos.pdf: 1081987 bytes, checksum: 435fc2cb438881b9c3905e16b4b41ed0 (MD5)
Made available in DSpace on 2016-08-03T14:07:42Z (GMT). No. of bitstreams: 1 Luiz Jose Rodrigues dos Santos.pdf: 1081987 bytes, checksum: 435fc2cb438881b9c3905e16b4b41ed0 (MD5) Previous issue date: 2008-02-28
Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES
Urban traffic represents a phenomenon of great socioeconomic importance,whose modeling from the point of view of prevision on the basis of initial conditions, still represents a challenge for modern science. Computational methods (computer simulations) represent a powerful tool for modeling and prediction of a number of effects, where systems of coupled differential equations may be used to simulate different phenomena observed in traffic systems. In particular, a quantity of high importance for maintenance and planning of road systems is the vehicular capacity which can be supported without traffic jams, whose description and prevision is still not well understood. In this work, a generalization of an existing microscopic traffic model, the Intelligent Driver Model (IDM), is proposed by implementing a distribution of desired velocities, where it is shown that vehicle capacity of multiple lane roads can be measured in a rather realistic manner, as a function of model parameters,which may be adjusted to real observations.
O tráfego urbano representa um fenômeno de grande importância sócio econômica, cuja modelagem de ponto de vista de previsão a partir de condições iniciais, ainda representa um desafio para a ciência moderna. Métodos computacionais (simulação computacional) representam uma ferramenta poderosa para modelagem e previsão de diversos efeitos, nos quais sistemas de equações diferenciais acopladas podem simular diversos fenômenos observados no sistema de tráfego. Em particular, uma grandeza de alto impacto para o gerenciamento e planejamento de rodovias é a capacidade veicular que elas podem suportar sem que aconteça o efeito de congestionamento, cuja descrição e previsão ainda não estão bem entendida. Neste trabalho, propõe-se uma generalização de um modelo microscópico computacional existente, o Intelligent Driver Model (IDM), aplicando uma distribuição de velocidades desejadas, onde torna-se possível medir de forma bastante realista a capacidade veicular de rodovias com múltiplas faixas, em função de parâmetros de modelo, que podem ser ajustados às observações reais.

As the field of autonomous robotics grows and its applications broaden up, an enormous amount of sensors and actuators, sometimes redundant, have been added to mobile robots. These now fully equipped entities are expected to perceive and act in their surrounding world in a human-like fashion, through perception, reasoning, planning and decision making processes. The higher complexity level of the resulting system and the nature of the environments where autonomous robots are usually expected to operate - continuous, partially unknown and usually unpredictable - demand the application of techniques to deal with this overload of data. In humans, that face the same problem when sounds, images and smells are presented to their sensors in a daily scene, a natural filter is applied: Attention. Although there are many computational models that apply attentive systems to Robotics, they usually are restricted to two classes of systems: a) those that have complex biologically-based attentional visual systems and b) those that have simpler attentional mechanisms with a larger variety of sensors. In order to evaluate an attentional system that operates with other robotics sensors than visual ones, this work presents a biologically inspired computational attentional model that can handle both top-down and bottom-up attention and that is able to learn how to re-distribute its limited resources over time and space. Experiments performed on a high fidelity simulator demonstrates the feasibility of the proposed attentional model and its capability on performing decision making and learning processes over attentional modulated data. The proposed system promotes a significant reduction on the original state space (96%) that was created over multiple sensory systems.

Knowledge intensive applications require an intelligent environment, which can perform deductions in response to user queries or events that occur inside or outside of the applications. For that, we propose a Fuzzy Petri Net (FPN) model to represent the knowledge and the behavior in an intelligent object-oriented database environment, which integrates fuzzy, active and deductive rules with database objects. By gaining intelligent behaviour, the system maintains objects to perceive dynamic occurences and user queries. Thus, objects can produce new knowledge or keep themselves in a consistent, stable, and upto-date state. The behavior of a system can be unpredictable due to the rules triggering or untriggering each other (non-termination). Intermediate and final database states may also differ according to the order of rule executions (non-confluence). In order to foresee and solve problematic behavior patterns, we employ static rule analysis on the FPN structure that provides easy checking of the termination property without requiring any extra construct. In addition, with our proposed inference algorithm, we guarantee confluent rule executions. The techniques and solutions provided in this study can be utilized in various complex systems, such as weather forecasting applications, environmental information systems, defense applications, video database applications, etc. We implement a prototype of the model for the weather forecasting of the Central Anatolia Region

With ever more enterprises holding most of their text documents electronically, this research considers the implications that follow and the opportunities that are created by this situation. It introduces the notion of intelligent document management through enrichment as a novel interpretation of the Push Technology paradigm and develops a conceptual model for a software assistant that can carry out this enrichment task. Such an Intelligent Document Enriching Assistant, IDEA, will perform simple reasoning to retrieve, automatically, previously stored text documents which are relevant to the task at hand. However, before such an assistant can actually be implemented, the underlying concepts, which are taken from a range of topics on which such an IDEA is founded, have to be researched and assembled. This is what is of concern here, namely the analysis of the background on which the IDEA is based, together with the development of its conceptual model. The implementation of this knowledge-based component does not form part of this work. The semantic content of the text document needs to be captured using a representation that is richer than simple key words, and is based on Speech Act Theory. For the development of the conceptual model for the IDEA, the CommonKADS method is used. The main deliverable of this method is the model of expertise, which describes the knowledge, and the reasoning capabilities needed to carry out the required enrichment task. The generic enrichment problem solving method is also developed. The underlying domain knowledge is taken from a Sales and Marketing scenario. This research answers the call for good practice in Knowledge Management that requires the development of new knowledge, making accessible new and existing knowledge, and distributing and combining knowledge. The IDEA is an example of how intelligence, through the provision of relevant information, can be realised within this context.

Thesis (M.S.)--Worcester Polytechnic Institute.
Keywords: cognitive model; model-tracing; intelligent tutoring system; tutoring; artificial intelligence. Includes bibliographical references (p. 55-57).

International Telemetering Conference Proceedings / October 21, 2002 / Town & Country Hotel and Conference Center, San Diego, California
Embedded GIS in Intelligent Navigation System is a special information system. This paper puts forward several basic principles and constraints during design for Embedded GIS at first, and then analyzes the feature of embedded platform and the function of Intelligent Navigation System, and presents a realization scheme of Embedded GIS.

Contact Centres (CCs) are at the forefront of interaction between an organisation and its customers. Currently, 17 percent of all inbound calls are not resolved on the first call by the first agent attending to that call. This is due to the inability of the contact centre agents (CCAs) to diagnose customer queries and find adequate solutions in an effective and efficient manner. The aim of this research is to develop an intelligent user interface (IUI) model to support and improve CC operations. A literature review of existing IUI architectures, modelbased design and existing CC software together with a field study of CCs has resulted in the design of an IUI model for CCs. The proposed IUI model is described in terms of its architecture, component-level design and interface design. An IUI prototype has been developed as a proof of concept of the proposed IUI model. The IUI prototype was evaluated in order to determine to what extent it supports problem identification and query resolution. User testing, incorporating the use of eye tracking and a post-test questionnaire, was used in order to determine the usability and usefulness of the prototype. The results of this evaluation show that the users were highly satisfied with the task support and query resolution assistance provided by the IUI prototype. This research resulted in the design of an IUI model for the domain of CCs. This model can be used to assist the development of CC applications incorporating IUIs. Use of the proposed IUI model is expected to support and enhance the effectiveness and efficiency of CC operations. Further research is needed to conduct a longitudinal study to determine the impact of IUIs in the CC domain.

This thesis develops and evaluates an intelligent model predictive control (IMPC) strategy for motion control of a flexible link robotic manipulator through analysis, computer simulation, and physical experimentation. The developed IMPC is based on a two-level hierarchical control architecture. This control structure is used to combine the advantages of the conventional model predictive control (MPC) with those of knowledge-based soft control techniques. The upper level of the structure is a fuzzy-rule based intelligent decision-making system. The lower level consists of two modules: a real-time system identification module (which adjusts the model parameters and accommodates payload changes of the manipulator), and a model predictive control (MPC) module (which develops control inputs based on the linear model generated by the system identification module). The upper-level intelligent fuzzy rule-based tuner interacts with the lower level modules. Based on the desired system performance, the state feedback signals, and the knowledge base, the upper-level fuzzy tuner automatically adjusts the tuning parameters of the MPC controller. It is also able to adjust the model structure of the system-identification module, if necessary, to accommodate large model errors, and will increase the robustness of the controller. An explicit, complete, and accurate nonlinear dynamic model of the system is developed using the assumed mode method. More realistic boundary conditions, which represent the balance of moments and shear forces separately, at the ends of each link, are used for the dynamic model development of the system. A computationally efficient multi-stage MPC algorithm with guaranteed stability is developed as well. This algorithm is used by the MPC module to enable real-time implementation of the overall scheme. A fuzzy knowledge base for tuning the MPC controller is developed based on analysis, computer simulations and experimental testing of the prototype flexible-link manipulator system (FLMS). A fuzzy tuner is designed based on this fuzzy knowledge base. The performance of the developed IMPC scheme is evaluated using computer simulations and experiments of the prototype FLMS. The results show that IMPC can more effectively control the motion of a flexible link robot manipulator when compared with conventional MPC.
Applied Science, Faculty of
Mechanical Engineering, Department of
Graduate

Complex under-actuated multilink mechanism involves a system whose number of control inputs is smaller than the dimension of the configuration space. The ability to control such a system through the manipulation of its natural dynamics would allow for the design of more energy-efficient machines with the ability to achieve smooth motions similar to those found in the natural world. This research aims to understand the complex nature of the Robogymnast, a triple link underactuated pendulum built at Cardiff University with the purpose of studying the behaviour of non-linear systems and understanding the challenges in developing its control system. A mathematical model of the robot was derived from the Euler-Lagrange equations. The design of the control system was based on the discrete-time linear model around the downward position and a sampling time of 2.5 milliseconds. Firstly, Invasive Weed Optimization (IWO) was used to optimize the swing-up motion of the robot by determining the optimum values of parameters that control the input signals of the Robogymnast’s two motors. The values obtained from IWO were then applied to both simulation and experiment. The results showed that the swing-up motion of the Robogymnast from the stable downward position to the inverted configuration to be successfully achieved. Secondly, due to the complex nature and nonlinearity of the Robogymnast, a novel approach of modelling the Robogymnast using a multi-layered Elman neural ii network (ENN) was proposed. The ENN model was then tested with various inputs and its output were analysed. The results showed that the ENN model to be capable of providing a better representation of the actual system compared to the mathematical model. Thirdly, IWO is used to investigate the optimum Q values of the Linear Quadratic Regulator (LQR) for inverted balance control of the Robogymnast. IWO was used to obtain the optimal Q values required by the LQR to maintain the Robogymnast in an upright configuration. Two fitness criteria were investigated: cost function J and settling time T. A controller was developed using values obtained from each fitness criteria. The results showed that LQRT performed faster but LQRJ was capable of stabilizing the Robogymnast from larger deflection angles. Finally, fitness criteria J and T were used simultaneously to obtain the optimal Q values for the LQR. For this purpose, two multi-objective optimization methods based on the IWO, namely the Weighted Criteria Method IWO (WCMIWO) and the Fuzzy Logic IWO Hybrid (FLIWOH) were developed. Two LQR controllers were first developed using the parameters obtained from the two optimization methods. The same process was then repeated with disturbance applied to the Robogymnast states to develop another two LQR controllers. The response of the controllers was then tested in different scenarios using simulation and their performance was evaluated. The results showed that all four controllers were able to balance the Robogymnast with the fastest settling time achieved by WMCIWO with disturbance followed by in the ascending order: FLIWOH with disturbance, FLIWOH, and WCMIWO.

High-speed trains have become one of the main means of public transportation around the world. The use of tilting train technologies on high-speed trains has contributed to cost effectiveness by reducing journey time between two places without the need to develop a new high-speed rail track infrastructure. Current technologies in tilting railway vehicles use a 'precedence' control scheme. This scheme uses a measurement from the front vehicle to capture 'precedence' information. Research on local sensor loop control strategies is still important to overcome the complexity of using precedence control technique. Work using conventional and modern control approaches has been investigated by previous researches. This study further extends these by investigating a particular intelligent control technique using fuzzy logic in designing the local feedback tilt control scheme.

The aim of this study is to understand the complexity and control challenges of the locomotion of a three-link mechanism of a robot system. In order to do this a three-link robot gymnast (Robogymnast) has been built in Cardiff University. The Robogymnast is composed of three links (one arm, one torso, one leg) and is powered by two geared DC motors. Currently the robot has three potentiometers to measure the relative angles between adjacent links and only one tachometer to measure the relative angular position of the first link. A mathematical model for the robot is derived using Lagrange equations. Since the model is inherently nonlinear and multivariate, it presents more challenges when modelling the Robogymnast and dealing with control motion problems. The proposed approach for dealing with the design of the control system is based on a discrete-time linear model around the upright position of the Robogymnast. To study the swinging motion of the Robogymnast, a new technique is proposed to manipulate the frequency and the amplitude of the sinusoidal signals as a means of controlling the motors. Due to the many combinations of the frequency and amplitude, an optimisation method is required to find the optimal set. The Bees Algorithm (BA), a novel swarm-based optimisation technique, is used to enhance the performance of the swinging motion through optimisation of the manipulated parameters of the control actions. The time taken to reach the upright position at its best is 128 seconds. Two different control methods are adopted to study the balancing/stablising of the Robogymnast in both the downward and upright configurations. The first is the optimal control algorithm using the Linear Quadratic Regulator (LQR) technique with integrators to help achieve and maintain the set of reference trajectories. The second is a combination of Local Control (LC) and LQR. Each controller is implemented via reduced order state observer to estimate the unmeasured states in terms of their relative angular velocities. From the identified data in the relative angular positions of the upright balancing control, it is reported that the maximum amplitude of the deviation in the relative angles on average are approximately 7.5° for the first link and 18° for the second link. It is noted that the third link deviated approximately by 2.5° using only the LQR controller, and no significant deviation when using the LQR with LC. To explore the combination between swinging and balancing motions, a switching mechanism between swinging and balancing algorithm is proposed. This is achieved by dividing the controller into three stages. The first stage is the swinging control, the next stage is the transition control which is accomplished using the Independent Joint Control (IJC) technique and finally balancing control is achieved by the LQR. The duration time of the transition controller to track the reference trajectory of the Robogymnast at its best is found to be within 0.4 seconds. An external disturbance is applied to each link of the Robogymnast separately in order to study the controller's ability to overcome the disturbance and to study the controller response. The simulation of the Robogymnast and experimental realization of the controllers are implemented using MATLAB® software and the C++ program environment respectively.

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

A aprendizagem assistida por computador, conhecida mais amplamente com o nome genérico de e-learning, converteu-se numa poderosa ferramenta com amplas potencialidades dentro do campo educativo. Mesmo assim, uma das maiores críticas que esta recebe é que na maioria dos casos os cursos que são implementados seguem um enfoque “one size fits all”, isto é, que todos os alunos recebem exatamente o mesmo conteúdo e da mesma maneira desconhecendo suas necessidades particulares. Esta falha radica não só na falta de interação direita entre aluno e tutor, senão também na falta de um desenho instrucional apropriado que considere alguns dos diversos enfoques disponíveis hoje em dia. Existem diversos enfoques que procuram solucionar este problema e adaptar o processo de ensino os estudantes. Pode-se dizer que na vanguarda de estes enfoques encontram-se os Sistemas Educacionais Inteligentes Adaptativos, os quais combinam as funcionalidades de dois enfoques: os Sistemas Hipermídia Educacionais Adaptativos y os Sistemas Tutoriais Inteligentes. Embora, logo de uma extensa revisão bibliográfica, se encontrou que existe ainda um inconveniente importante com este tipo de sistemas e em particular com seus modelos de referência: ou são demasiado simples, incluindo somente umas poucas funcionalidades; ou são demasiado complexos, o que dificulta seu desenho e implementação. Considerando este panorama, o objetivo principal de esta tese foi a definição de um modelo de referência intentando alcançar esse equilíbrio esquivo, de tal maneira que permita o desenho de cursos que se adaptem de una maneira efetiva e inteligente ao progresso e características de cada estudante, mas sem ser demasiado complexo. Outra propriedade importante desse modelo és que integra o uso de Objetos de Aprendizagem, promovendo assim a flexibilidade e a usabilidade. Para alcançar este objetivo geral, três sub modelos foram considerados: um modelo do domínio, um modelo do estudante y um modelo do tutor. O primeiro serve para estruturar o domínio de conhecimento e foi definido usando a noção de objetivo de aprendizagem junto com um esquema flexível multi-nível com operações opcionais de pré-requisitos. O segundo visa caracterizar aos estudantes e considera informação pessoal, de conhecimento e psico-cognitiva. O terceiro pode ser considerado como o coração do sistema e define as funcionalidades adaptativas consideradas: sequenciamento y navegação, apresentação de conteúdo, evacuação, y suporte colaborativo. Com o fim de clarificar os três sub modelos, assim como todos seus componentes e relações, se presentou um exemplo de instanciação que se denominou Doctus, o qual consiste em una ferramenta de autor para cursos adaptativos. Doctus não somente serviu para exemplificar o uso do modelo de referência em sua totalidade, mas também para refinar os sub modelos e alguns procedimentos involucrados. Como parte final de esta tese, se realizou também a implementação e validação preliminar de Doctus. Isto foi feito com 51 sujeitos, professores em diversos níveis de formação. Os resultados obtidos em esta etapa foram sobressalientes no sentido que todas as funcionalidades adaptativas foram bem avaliadas e todos os pesquisados manifestaram seu entusiasmo por contar com uma ferramenta que lhes ajudara em seus práticas docentes considerando a seus estudantes como indivíduos particulares.
Computer Aided Learning, known more widely with the generic name of e-learning, has become a powerful tool with lots of potentialities within educational field. Even though, one of the main critics that it receives is that in most cases the implemented courses follows a “one size fits all” approach, which means that all students receive the same content in the same way being unaware of their particular needs. This problem is not due only to the absence of direct interaction between student and tutor, but also because of the lack of an appropriate instructional design. There are several approaches which deal with this issue and look for adapt the teaching process to students. One could say that in the top of those approaches the Adaptive and Intelligent Educational Systems are situated, which merges the functionalities of two approaches: the Adaptive Educational Hypermedia Systems and the Intelligent Tutoring Systems. Nevertheless, after an extensive literature review, a major inconvenience is still found for this kind of systems and particularly for their reference models: or they are too simple, including just a few functionalities; or they are too complex, which difficult their design and implementation. Considering this panorama, the main objective of this dissertation thesis was the definition of a reference model trying to reach such an elusive equilibrium, in such a way that allows the design of courses which adapt themselves in an intelligent and effective way to the progress and characteristics of each student but without being too complex. Another important feature is that this model integrates Learning Objects, promoting this way flexibility and reusability. In order to achieve this general objective, three sub-models were considered: a domain model, a student model and a tutor model. The first one serves to structure the knowledge domain and was defined using the notion of learning goal and a flexible multilevel schema with optional prerequisite operations. The second one aids to characterize students and considered personal, knowledge and psycho-cognitive information. The third one may be considered as the hearth of the system and defines the adopted adaptive functionalities: sequencing and navigation, content presentation, assessment, and collaborative support. With the aim of clarify the three sub-models, as well as all their components and relationships, an instantiation example was also presented. Such an instantiation was called Doctus, an authoring tool for adaptive courses. Doctus was not only helpful to exemplify the setup of the referece model as a whole, but also to refine sub-models and several procedures envolved. As final part of the dissertation, the implementation and preliminary validation of Doctus was performed. This was done with 51 subjects, teachers from different formation levels. The obtained results in this stage were outstanding, all the adaptive functionalities were well evaluated and all of those polled felt enthusiastic about counting with a tool for helping them in their teaching practices considering students as particular individuals.
El aprendizaje asistido por computador, conocido más ampliamente con el nombre genérico de e-learning, se ha convertido en una poderosa herramienta con amplias potencialidades dentro del campo educativo. Aun así, una de las mayores críticas que este recibe es que en la mayoría de los casos los cursos que son implementados siguen un enfoque “one size fits all”, es decir, que todos los alumnos reciben exactamente el mismo contenido y de la misma manera desconociendo sus necesidades particulares. Esta falla radica no sólo en la falta de interacción directa entre alumno y tutor, sino también en la falta de un diseño instruccional apropiado que considere diversos de los enfoques disponibles hoy en día. Existen diversos enfoques que buscan solucionar este problema y adaptar el proceso de enseñanza a los estudiantes. Se podría decir que a la vanguardia de estos enfoques se encuentran los Sistemas Educacionales Inteligentes Adaptativos, los cuales combinan las funcionalidades de dos enfoques: los Sistemas Hipermedia Educacionales Adaptativos y los Sistemas Tutoriales Inteligentes. Sin embargo, luego de una extensa revisión bibliográfica, se encontró que existe aún un inconveniente importante con este tipo de sistemas y en particular con sus modelos de referencia: o son demasiado simples, incluyendo solamente unas pocas funcionalidades; o son demasiado complejos, lo cual dificulta su diseño e implementación. Considerando este panorama, el objetivo principal de esta tesis fue la definición de un modelo de referencia intentando alcanzar tal equilibrio esquivo, de tal manera que permita el diseño de cursos que se adapten de una manera efectiva e inteligente al progreso y características de cada estudiante pero sin ser demasiado complejo. Otra propiedad importante de dicho modelo es que integra el uso de Objetos de Aprendizaje, promoviendo así la flexibilidad y la reusabilidad. Con el fin de alcanzar este objetivo general, tres sub modelos fueron considerados: un modelo del dominio, un modelo del estudiante y un modelo del tutor. El primero sirve para estructurar el dominio de conocimiento y fue definido empleando la noción de objetivo de aprendizaje junto con un esquema flexible multinivel con operaciones opcionales de prerrequisitos. El segundo busca caracterizar los estudiantes y considera información personal, de conocimiento y psico-cognitiva. El tercero puede ser considerado como el corazón del sistema y define las funcionalidades adaptativas consideradas: secuenciamiento y navegación, presentación de contenido, evaluación, y soporte colaborativo. Con el fin de clarificar los tres sub modelos, así como todos sus componentes y relaciones, se presentó además un ejemplo de instanciación. Tal instanciación se denominó Doctus, el cual consiste en una herramienta de autor para cursos adaptativos. Doctus no solamente sirvió para ejemplificar el uso del modelo de referencia en su totalidad, sino también para refinar los sub modelos y algunos procedimientos involucrados. Como parte final de esta tesis, se realizó también la implementación y validación preliminar de Doctus. Esto se hizo con 51 sujetos, todos profesores en diversos niveles de formación. Los resultados obtenidos en esta etapa fueron sobresalientes en el sentido que todas las funcionalidades adaptativas fueron bien evaluadas y todos los encuestados manifestaron su entusiasmo por contar con una herramienta que les ayudara en sus prácticas docentes considerando a sus estudiantes como individuos particulares.

This research work focuses at first on the intelligent model development for process state (special for fault) detection, behavior prediction and process control for complex industrial processes. In the model architecture, Fuzzy Neural Networks (FNNs) are employed as process state classifiers for process state (fault) detection; other (different) Neural Networks (NNs) models are applied for system identification of process characteristics in different process states. The model detects process states (faults) and predicts process behavior according to process input and historical behaviors, whose combination of influences generates the final results of process state (fault) detection and quantitative prediction. The whole model is constructed based on Fuzzy TS NARX models. Secondly, an optimal model is designed to two purposes, one is for optimal process diagnosis and another is for optimal prediction. To time varying processes, an adaptive strategy and algorithm, applying the Least Squares algorithm, has been developed for model adaptability to cover time depending process changes. Thirdly, a specific state space equation of discrete time varying system is being derived from the model. In the state space equation, the state transition matrix A is determined by the fuzzy degree of process state classification produced by process historical behavior in time t instant, and the input transition matrix B by process real input in time t instant. The state observer vector H is determined by optimization results generated by model adaptive or optimal scheme. Finally, to confirm the validity of the theoretical results from above, an application case has been studied for supply forecasting. The study and application results indicate that the model not only has good performance for fault detection, but also provides excellent quantitative prediction of process output. It can be applied in process state (fault) detection, diagnosis and prediction for process behavior, as well as fault predictive control.

This thesis focuses on two important problems related to DDDAS: interest management (data distribution) and prediction models. In order to reduce communication overhead, we propose a new interest management mechanism for mobile peer-to-peer systems. This approach involves dividing the entire space into cells and using an efficient sorting algorithm to sort the regions in each cell. A mobile landmarking scheme is introduced to implement this sort-based scheme in mobile peer-to-peer systems. The design does not require a centralized server, but rather, every peer can become a mobile landmark node to take a server-like role to sort and match the regions. Experimental results show that the scheme has better computational efficiency for both static and dynamic matching. In order to improve communication efficiency, we present a travel time prediction model based on boosting, an important machine learning technique, and combine boosting and neural network models to increase prediction accuracy. We also explore the relationship between the accuracy of travel time prediction and the frequency of traffic data collection with the long term goal of minimizing bandwidth consumption. Several different sets of experiments are used to evaluate the effectiveness of this model. The results show that the boosting neural network model outperforms other predictors.

Intelligent Tutoring Systems attempt to create a relationship between a computer and the student which resembles a human-to-human tutorial situation. For successful teaching to take place an Intelligent Tutoring System has to be able to cope with any student errors that may occur during a consultation. Remedial intervention implemented in current Intelligent Tutoring Systems lacks a formal basis. The objective of this research is to formalise the process of remediation with Intelligent Tutoring Systems and to provide a framework for the implementation of remedial tutoring in Intelligent Tutoring Systems. This research first presents a state-of-the-art account of Intelligent Tutoring Systems. It then proceeds with an investigation of both current practices with existing Intelligent Tutoring Systems and requirements for providing remedial tutoring. What stems from this investigation is a set of principles that governs remedial tutoring intervention. These principles of remediation serve as the foundation for the construction of the model for remedial operations, which can be employed in developing Intelligent Tutoring Systems capable of offering remedial tutoring. To demonstrate this, INTUITION, an Intelligent Tutoring System implementation of an existing business simulation game, is developed. The thesis then proposes an evaluation method which can be used to assess remedial intervention with Intelligent Tutoring Systems against the principles of remediation. This evaluation method is applied to INTUITION. The result of the evaluation shows that INTUITION follows the principles of remediation and that, therefore, the model for remedial operations is a useful method for providing remedial tutoring with Intelligent Tutoring Systems according to the principles of remediation.

Thesis: S.M. in Engineering and Management, Massachusetts Institute of Technology, System Design and Management Program, 2017.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 101-104).
With the recent resurgence of Machine Learning and Artificial Intelligence as a competitive advantage in product development, technical executives and managers are interested in learning what it would take to build intelligent platforms that can leverage these advances. In addition, they wish to produce cost estimates for developing such platforms. The goal of this thesis is to develop a reference architecture for an intelligent platform and an associated costing model that allows technical managers to understand the components needed to deliver such a platform and estimate the cost of each module, estimate the cost of the overall architecture, and enable what-if analysis to understand the cost tradeoffs. The intent is not to provide the values of the variables in the model, but to develop a cost model that will enable interested parties to plug in their estimated values for each factor and generate a forecast of the build cost.
by Mohammad Jouni.
S.M. in Engineering and Management

La subjectivité des estimations et des perceptions, la complexité de l’environnement, l’interaction entre sous-systèmes, le manque de données précises, les données manquantes, une faible capacité de traitement de l’information, et l’ambiguïté du langage naturel représentent les principales formes d’incertitude auxquelles les décideurs doivent faire face lorsqu’ils prennent des décisions stratégiques à l’aide de systèmes d’intelligence économique. Cette étude utilise un paradigme de « soft computing » pour identifier et analyser l’incertitude, que nous associons à la notion de facteurs de risque d’information. Pour cela, nous proposons un modèle de rapprochement exploitant des ontologies, ainsi qu’un modèle baptisé « FuzzyWatch » fondé sur la logique floue. Nous avons modélisé le processus de prise de décision depuis la définition du problème jusqu’à la réponse à la question : « est-il raisonnable de décider ? ». Un diagramme causal d’Ishikawa permet de prendre en compte les facteurs intangibles dans cette approche. Le cadre de référence du rapprochement de connaissances a été prévu pour faciliter le partage et la réutilisation de connaissances entre les utilisateurs et la machine. En complément, les facteurs intangibles, les émotions, les ambiguïtés du langage naturel sont pris en compte à l’aide de fonctions d’appartenance floues. Les outils de la logique floue ont été également utilisés au niveau des ontologies (« FuzzOntology »). Au niveau du processus de recherche d’information, l’introduction d’une fonction de mise en correspondance floue, appelée « FuzzyMatch », améliore le taux de rappel et subséquemment le processus d’intelligence économique. Le modèle « Fuzzontologique » autorise une prise en compte flexible de facteurs intangibles et incertains, offrant ainsi un moyen de traiter l’ambiguïté du langage naturel. FuzzyMatch permet de réduire les problèmes de données manquantes. A l’aide de ces modèles, le processus de décision en intelligence économique bénéficie d’une réduction des risques liés à l’information lors du processus de recherche
Subjective estimation and perception, complexity of the environment under study, interaction amongst subsystems, lack of precise data, missing data, limited information processing capacity and ambiguity in natural languages are major forms of uncertainty facing Decision Makers in the process of delivering strategic decisions in economic intelligent systems. This study employs soft computing paradigm to capture and analyze uncertainty based on information risk factors via our proposed knowledge reconciliation model based on ontology and the FuzzyWatch model. We modeled the process of decision making from the point of problem definition to decision delivery (translation credibility) and include intangible factors with the fish-bone architecture. Ontological framework for Knowledge Reconciliation was developed to facilitate knowledge sharing and reuse among both human and computer agents while intangible factors, emotions and ambiguities in natural languages were captured with fuzzy membership function. We extended this operation with fuzzy that is – what ontology captures is interpreted by fuzzy techniques (FuzzOntology). The fuzzy match relation for information retrieval tagged “FuzzyWatch” improves the information search result thus reducing the risk of missing data which is of grave consequence in Economic Intelligence process. FuzzOntological model facilitates a flexible means of capturing intangible and uncertain factors as a means of resolving the ambiguity in natural languages. FuzzyWatch assists in reducing missing data problems. Future decisional process will contend with lesser information retrieval risks in Economic Intelligence process using this model

Sensor-embedded tires, known as intelligent tires, have been widely studied because they are believed to provide reliable and crucial information on tire-road contact characteristics e.g., slip, forces and deformation of tires. Vehicle control systems such as ABS and VSP (Vehicle Stability Program) can be enhanced by leveraging this information since control algorithms can be updated based on directly measured parameters from intelligent tire rather than estimated parameters based on complex vehicle dynamics and on-board sensor measurements. Moreover, it is also expected that intelligent tires can be utilized for the purpose of the analysis of tire characteristics, taking into consideration that the measurements from the sensors inside the tire would contain considerable information on tire behavior in the real driving scenarios. In this study, estimation methods for the tire-road contact features by utilizing intelligent tires are investigated. Also, it was discussed how to identify key tire parameters based on the fusion technology of intelligent tire and tire modeling. To achieve goals, extensive literature reviews on the estimation methods using the intelligent tire system was conducted at first. Strain-based intelligent tires were introduced and tested in the laboratory for this research. Based on the literature review and test results, estimation methods for diverse tire-road contact characteristics such as slippages and contact forces have been proposed. These estimation methods can be grouped into two categories: statistical regressions and model based methods. For statistical regressions, synthetic regressors were proposed for the estimation of contact parameters such as contact lengths, rough contact shapes, test loads and slip angles. In the model-based method, the brush type tire model was incorporated into the estimation process to predict lateral forces. Estimated parameters using suggested methods agreed well with measured values in the laboratory environment. By utilizing sensor measurements from intelligent tires, the tire physical characteristics related to in-plane dynamics of the tire, such as stiffness of the belt and sidewall, contact pressure distribution and internal damping, were identified based on the combination of strain measurements and a flexible ring tire model. The radial deformation of the tread band was directly obtained from strain measurements based on the strain-deformation relationship. Tire parameters were identified by fitting the radial deformations from the flexible ring model to those derived from strain measurements. This approach removed the complex and repeated procedure to satisfy the contact3 constraints between the tread and the road surface in the traditional ring model. For tires with different specifications, identification using the suggested method was conducted and their results are compared with results from conventional methods and tests, which shows good agreements. This approach is available for the tire standing still or rolling at low speeds. For tires rolling at high speeds, advanced tire model was implemented and associated with strain measurements to estimate dynamic stiffness, internal damping effects as well as dynamic pressure distributions. Strains were measured for a specific tire under various test conditions to be used in suggested identification methods. After estimating key tire parameters step by step, dynamic pressure distributions was finally estimated and used to update the estimation algorithm for lateral forces. This updated estimation method predicted lateral forces more accurately than the conventional method. Overall, this research will serve as a stepping stone for developing a new generation of intelligent tire capable of monitoring physical tire characteristics as well as providing parameters for enhanced vehicle controls.
PHD

The goal of this research was to determine if the results of a simple comparison algorithm (SCA) could be improved by adding a hyperspace analogue to language model of memory (HAL) layer to form NCA. The HAL layer provides contextual data that otherwise would be unavailable for consideration. It was found that NCA did improve the results when compared to SCA alone. However, NCA added complexity problems that limit its practicality. The complexity of this algorithm is On3 where n is equal to the number of unique symbols in the data. While there is a relativity reasonable soft upper bound for the number of unique symbols used in a language, the complexity still limits the uses of the NCA combined algorithm. The conclusion from this research is that NCA can improve results. This research also suggested that the quality of results might increase as more data is processed by NCA.

This paper's main objective is to consolidate the knowledge on context in the realm of intelligent systems, systems that are aware of their context and can adapt their behavior accordingly. We provide an overview and analysis of 36 context models that are heterogeneous and scattered throughout multiple fields of research. In our analysis, we identify five shared context categories: social context, location, time, physical context, and user context. In addition, we compare the context models with the context elements considered in the discourse on intelligent systems and find that the models do not properly represent the identified set of 3,741 unique context elements. As a result, we propose a consolidation of the findings from the 36 context models and the 3,741 unique context elements. The analysis reveals that there is a long tail of context categories that are considered only sporadically in context models. However, particularly these context elements in the long tail may be necessary for improving intelligent systems' context awareness.

Adaptive tutorial dialogues have been successfully employed by ITSs to facilitate deep learning of conceptual domain knowledge. But none of the approaches used for generating dialogues have been used across instructional domains and tasks. The objective of this project was twofold: (i) to propose a general model that provides adaptive dialogue support in both well- and ill-defined instructional tasks (ii) to explore whether adaptive tutorial dialogues are better than non-adaptive dialogues in acquiring domain knowledge. Our model provides adaptive dialogue support by identifying the concepts that the student has most difficulty with, and then selecting the tutorial dialogues corresponding to those concepts. The dialogues are customised based on the student’s knowledge and explanation skills, in terms of the length and the exact content of the dialogue. The model consists of three parts: an error hierarchy, tutorial dialogues and rules for adapting them. We incorporated our model into EER-Tutor, a constraint-based tutor that teaches database design. The effectiveness of adaptive dialogues compared to non-adaptive dialogues in learning this ill-defined task was evaluated in an authentic classroom environment. The results revealed that the acquisition of the domain knowledge (represented as constraints) of the experimental group who received adaptive dialogues was significantly higher than their peers in the control group with non-adaptive dialogues. We also incorporated our model into NORMIT, a constraint-based tutor that teaches data normalization. We repeated the experiment using NORMIT in a real-world class room environment with a much smaller group of students (18 in NORMIT study vs 65 in EER-Tutor study) but did not find significant differences. We also investigated whether our model could support dialogues in logical database design and fraction addition using paper-based methods. Our evaluation studies and investigations on paper indicated that our model can provide adaptive support for both ill-and well-defined tasks associated with a well-defined domain theory. The results also indicated that adaptive dialogues are more effective than non-adaptive dialogues in teaching the ill-defined task of database design.

Approved for public release; distribution is unlimited
In 1998, the Marine Corps initiated the Integrated Logistics Capability (ILC) to specifically address issues related to Marine Corps logistics doctrine, policy, and processes. The purposes of ILC initiatives were to define, measure, and improve core logistics capabilities to meet the challenges of the new millennium and beyond. During a four-week Best Business Practices seminar hosted by Penn State, several key concepts were analyzed that would improve cost leverage, enhance the robustness of supplier relationships, and substitute technology and information for inventory. In an effort to successfully integrate these issues, a segmentation methodology or Quadrant Model approach was introduced. The Quadrant Model is essentially a two-by-two matrix that classifies products, services, and inventory items into four major cells. Each cell is further categorized by uniqueness and value, and each cell implies different approaches to managing inventory, supply chains, and vendor relationships. The objective of this thesis is to determine how Intelligent Agent (IA) technologies may innovate the logistical processes associated the critical quadrant, which categorizes materials, products and services that are unique, high value in terms of cost, and therefore high risk. These materials may only be available from a few sources, are in limited quantities, and require extensive coordination and sharing of information between parties. The researcher applies a model derived from previous thesis work that examined the application of IA technology as a method to innovate the Standard Procurement System (SPS). Using a high-level process model of the "critical" quadrant, the researcher identifies two applications of IA technology that may support attributes within the "critical" quadrant.

Large-scale complex systems have large numbers of variables, network structure of interconnected subsystems, nonlinearity, spatial distribution with several time scales in its dynamics, uncertainties and constrained. Decomposition of large-scale complex systems into smaller more manageable subsystems allowed for implementing distributed control and coordinations mechanisms. This thesis proposed the use of distributed softly switched robustly feasible model predictive controllers (DSSRFMPC) for the control of large-scale complex systems. Each DSSRFMPC is made up of reconfigurable robustly feasible model predictive controllers (RRFMPC) to adapt to different operational states or fault scenarios of the plant. RRFMPC reconfiguration to adapt to different operational states of the plant is achieved using the soft switching method between the RRFMPC controllers. The RRFMPC is designed by utilizing the off-line safety zones and the robustly feasible invariant sets in the state space which are established off-line using Karush Kuhn Tucker conditions. This is used to achieve robust feasibility and recursive feasibility for the RRFMPC under different operational states of the plant. The feasible adaptive cooperation among DSSRFMPC agents under different operational states are proposed. The proposed methodology is verified by applying it to a simulated benchmark drinking water distribution systems (DWDS) water quality control.

Variation in tutoring strategy plays an important part in Intelligent Tutoring Systems (ITSs). The potential for providing an adaptive ITS depends initially on having a range of tutoring strategies to select from. However, in order to react effectively to the student's needs, an ITS not only has to be able to simply offer different tutoring strategies but to choose intelligently among them and determine which one is best for an individual student at a particular moment. This thesis first examines, through literature review and interactions with existing systems, the current practices of ITSs regarding the provision of multiple tutoring strategies and tutoring strategy selection. What stems from this examination are the principles that underlie tutoring strategys election. These principles of tutoring strategy selection serve as a foundation for the construction of the model for tutoring strategy selection. To demonstrate the benefits of having such a model for formalising selection, the model is then implemented in ARISTOTLE, an existing ITS for tutoring zoology that includes several tutoring strategies but uses ad hoc mechanisms for choosing among them. This research is therefore contributing, through the principles of, and the model for tutoring strategy selection, a formal basis for selecting among tutoring strategies in ITSs that incorporate multiple tutoring strategies.

First the thesis reviews the development of Intelligent Computer Assisted Instruction (ICAI) systems by outlining the different ways that computers have been used in education followed by a description of the functionality of ICAI systems in terms of the Hartley-Sleeman model of classification. This is followed by a discussion of the skills required within writing and their pedagogical context. The different strategies that have been applied to computer supported composition are then discussed with examples of systems where appropriate.

The networks of surveillance cameras installed in public places and private territories continuously record video data with the aim of detecting and preventing unlawful activities. This enhances the importance of video content analysis applications, either for real time (i.e. analytic) or post-event (i.e. forensic) analysis. In this thesis, the primary focus is on four key aspects of video content analysis, namely; 1. Moving object detection and recognition, 2. Correction of colours in the video frames and recognition of colours of moving objects, 3. Make and model recognition of vehicles and identification of their type, 4. Detection and recognition of text information in outdoor scenes. To address the first issue, a framework is presented in the first part of the thesis that efficiently detects and recognizes moving objects in videos. The framework targets the problem of object detection in the presence of complex background. The object detection part of the framework relies on background modelling technique and a novel post processing step where the contours of the foreground regions (i.e. moving object) are refined by the classification of edge segments as belonging either to the background or to the foreground region. Further, a novel feature descriptor is devised for the classification of moving objects into humans, vehicles and background. The proposed feature descriptor captures the texture information present in the silhouette of foreground objects. To address the second issue, a framework for the correction and recognition of true colours of objects in videos is presented with novel noise reduction, colour enhancement and colour recognition stages. The colour recognition stage makes use of temporal information to reliably recognize the true colours of moving objects in multiple frames. The proposed framework is specifically designed to perform robustly on videos that have poor quality because of surrounding illumination, camera sensor imperfection and artefacts due to high compression. In the third part of the thesis, a framework for vehicle make and model recognition and type identification is presented. As a part of this work, a novel feature representation technique for distinctive representation of vehicle images has emerged. The feature representation technique uses dense feature description and mid-level feature encoding scheme to capture the texture in the frontal view of the vehicles. The proposed method is insensitive to minor in-plane rotation and skew within the image. The capability of the proposed framework can be enhanced to any number of vehicle classes without re-training. Another important contribution of this work is the publication of a comprehensive up to date dataset of vehicle images to support future research in this domain. The problem of text detection and recognition in images is addressed in the last part of the thesis. A novel technique is proposed that exploits the colour information in the image for the identification of text regions. Apart from detection, the colour information is also used to segment characters from the words. The recognition of identified characters is performed using shape features and supervised learning. Finally, a lexicon based alignment procedure is adopted to finalize the recognition of strings present in word images. Extensive experiments have been conducted on benchmark datasets to analyse the performance of proposed algorithms. The results show that the proposed moving object detection and recognition technique superseded well-know baseline techniques. The proposed framework for the correction and recognition of object colours in video frames achieved all the aforementioned goals. The performance analysis of the vehicle make and model recognition framework on multiple datasets has shown the strength and reliability of the technique when used within various scenarios. Finally, the experimental results for the text detection and recognition framework on benchmark datasets have revealed the potential of the proposed scheme for accurate detection and recognition of text in the wild.

In recent times, the Internet revolution has spawned numerous innovative enterprises-virtual companies, and electronic markets. Electronic markets (or digital markets) are scalable web-based platforms for buyers, sellers, marketmakers, and brokers to carry out business transactions. Over the last two years, there has been a proliferation of such E-Markets on the web. In this thesis, we develop an E-marketplace, which we call IMAGINE (Intelligent Market with AGents and Integrative NEgotiations) that improves upon the existing state-of-the-art in several non-trivial ways. IMAGINE combines the best features of existing E-marketplaces with several innovations. The thesis describes the conceptualization, analysis, and design of IMAGINE and provides details of implementation of a prototype of IMAGINE at the Electronic Enterprises Laboratory, Department of Computer Science and Automation, Indian Institute of Science. IMAGINE is a collaborative, co-operative, intelligent E-Market that maximizes the combined utility value of the all traders involved. IMAGINE has several distinctive features: • It uses an innovative business model, which is intelligent in the sense of perceiving the nature of the market and market forces and using this market intelligence in matching buyers with sellers and in determining the prices. • It uses integrative negotiations, which make it attractive for buyers and sellers to reveal their true business interests and valuations. • A sound and robust software architecture for a web-based implementation using best practices in object technology. • Implementation of a prototype of IMAGINE has been carried out using leading edge Internet technologies such as multi-agent technology, Jini, and Javaspaces.

Environment through the application of the established Model of Intelligent Assisted Built Residential Environment (IABRE). This model is used to establish the Multi-criteria Decision Support System of Intelligent Assisted Built Residential Environment (MDSSIABRE), in which multi-criteria analysis and multi-criteria alternative design methods are used. In order to reach this aim, the following tasks must be solved: • Established models and intelligent systems are analysed in various countries. The experience of Lithuania and foreign countries is analysed in the establishment of ambient assisted living environment through the application of modern internet, intelligent and other technologies and innovations; • A Model of Intelligent Assisted Built Residential Environment is estab-lished (MIABRE); • The system of criteria is established, which describes IABRE in detail; • Data base is established, which describes IABRE in detail; • MDSSIABRE system is established on the basis of the Model of Intelli-gent Assisted Built Residential Environment; • Practical realisation of the established IABRE system is performed. The thesis is divided into four chapters including the conclusion’s chapter. Chapter 1 analyses assisted residential environment, reviews the research on life quality and conditions in Lithuanian and abroad. It analyses the premises and the opportunities for creation of built and humanised environment. A review of current housing issues and housing development in... [to full text]
The aim of the thesis is the evaluation and enhancement of the effectiveness of Intelligent Assisted Built Residential Pirmajame disertacijojos skyriuje analizuojama gyventi palanki būsto aplinka, „intelektinio būsto“, „sumanaus namo“ sampratos, nagrinėjamos intelektine gyventi palankia aplinka suinteresuotos grupės, esamos būsto problemos Lietu-voje ir užsienyje. Aptariama kompiuterinių technologijų, nanotechnologijų įtaka gyventi palankios aplinkos kūrimui. Akcentuojamas intelektinis aplinką tausojantis būstas, jo įgyvendinimo būdai ir priemonės. Apibendrinami atlikti moksliniai tyrinėjimai intelektinio būsto kūrimo srityje, nagrinėjami gyventi palankios ap-linkos analizės modeliai. Antrajame skyriuje nagrinėjama integruotos intelektinės gyventi palankios aplinkos būsto modelio koncepcija. Pateikiamas ir detaliai analizuojamas integruoto intelektinės gyventi palankios aplinkos būsto modelis, nagrinėjami modelį sudarantys elementai, mikro- ir makroaplinka, jos įtaka, paklausos ir pasiūlos charakteristikos, praktinės modelio realizavimo galimybės. Trečiajame skyriuje analizuojamas intelektinės gyvenamosios aplinkos variantinis projektavimas ir daugiakriterinė analizė. Apibendrinamas daugiakriterinei analizei reikalingų pradinių duomenų paruošimo procesas. Aprašomas autorės pasiūlytas intelektinės gyvenamosios aplinkos variantinio projektavimo ir daugiakriterinės analizės modelis, analizės atlikimo eiliškumas, pateikiamos skaičiuoti reikalingos formulės. Atliekamas praktinių... [toliau žr. visą tekstą]

The widely used vector model maintains its popularity because of its simplicity, fast speed, and the appeal of using spatial proximity for semantic proximity. However, this model faces a disadvantage that is associated with the vagueness from keywords overlapping. Efforts have been made to improve the vector model. The research on improving document representation has been focused on four areas, namely, statistical co-occurrence of related items, forming term phrases, grouping of related words, and representing the content of documents. In this thesis, we propose the idea-indexing model to improve document representation for the filtering task in IR. The idea-indexing model matches document terms with the ideas they express and indexes the document with these ideas. This indexing scheme represents the document with its semantics instead of sets of independent terms. We show in this thesis that indexing with ideas leads to better performance.

The aim of this study was to understand, clarify, and explain emotional intelligence, as well as possible sub concepts that are correlated to, and might also influence and shape, the leadership style that comes from emotional intelligence. Through Daniel Goleman’s (1995) bestseller: Emotional Intelligence, the topic has emerged rapidly during the last two decades and has been implemented in various settings, and among them as a leadership style. Now, even though emotionally intelligent leadership (EIL) has been touched upon in research to a certain level, there is no established map that explains the leadership style or its context. That is the gap this thesis is trying to close. By analyzing five books (from representatives of both the ability model and the mixed model) concerning strictly emotionally intelligent leadership, I wanted to discover what contextual perception these authors had of the leadership style and its context. The empirical study showed that the authors discussed a total of nine sub concepts which could be placed in the surrounding of EIL. After the analysis, the nine concepts were narrowed down to seven, which shape and influence this leadership style. Those seven concepts are acting, communication, gender, humor, mood, symbolic management and training and development. The development of emotional intelligence, emotionally intelligent leadership and its contextual concepts have also been presented through a model called: Linnaeus Model of Emotionally Intelligent Leadership and its Contextual Sub Concepts. This paper is a first effort at mapping the terrain of this emotional way of leading, and I hope that scholars and researchers within the study area, as well as practitioners and consultants (especially within leadership, management and HR) will find this paper useful.

This thesis, first of all, proposes a new data mining paradigm (transaction-filtering association rule mining) addressing a time consumption issue caused by the repeated scans of original transaction databases in conventional associate rule mining algorithms. An in-memory transaction filter is designed to discard those infrequent items in the pruning steps. This filter is a data structure to be updated at the end of each iteration. The results based on an IBM benchmark show that an execution time reduction of 10% - 19% is achieved compared with the base case. Next, a data mining-based predictive model is then established contributing to intelligent data management within the context of Centre for Grid Computing. The capability of discovering unseen rules, patterns and correlations enables data mining techniques favourable in areas where massive amounts of data are generated. The past behaviours of two typical scenarios (network file systems and Data Grids) have been analyzed to build the model. The future popularity of files can be forecasted with an accuracy of 90% by deploying the above predictor based on the given real system traces. A further step towards intelligent policy design is achieved by analyzing the prediction results of files’ future popularity. The real system trace-based simulations have shown improvements of 2-4 times in terms of data response time in network file system scenario and 24% mean job time reduction in Data Grids compared with conventional cases.