Neurofuzzy design and model construction of nonlinear dynamical processes from data

A common problem in many data based modelling algorithms such as associative memory networks is the problem of the curse of dimensionality. In this paper, a new two-stage neurofuzzy system design and construction algorithm (NeuDeC) for nonlinear dynamical processes is introduced to effectively tackle this problem. A new simple preprocessing method is initially derived and applied to reduce the rule base, followed by a fine model detection process based on the reduced rule set by using forward orthogonal least squares model structure detection. In both stages, new A-optimality experimental design-based criteria we used. In the preprocessing stage, a lower bound of the A-optimality design criterion is derived and applied as a subset selection metric, but in the later stage, the A-optimality design criterion is incorporated into a new composite cost function that minimises model prediction error as well as penalises the model parameter variance. The utilisation of NeuDeC leads to unbiased model parameters with low parameter variance and the additional benefit of a parsimonious model structure. Numerical examples are included to demonstrate the effectiveness of this new modelling approach for high dimensional inputs.

Estruturas de governança corporativa e financial distress: há relação entre conselho de administração e empresas em financial distress?

Nesta dissertação foi analisada se há uma relação significante entre estruturas de governança (estrutura e composição de conselho) e financial distress. Este trabalho focou neste tema porque os estudos acadêmicos em governança corporativa e sua relação com financial distress ainda são pouco explorados. Além disso, o tema tem relevância no mundo corporativo, pois entender quais estruturas e composições de conselho seriam mais eficientes para evitar financial distress é interessante para diversos stakeholders, principalmente para os acionistas e os credores. Para verificar a existência dessa relação, foram utilizados dados de empresas brasileiras de capital aberto e foram desenvolvidos modelos logit de financial distress. Sendo a variável resposta financial distress, partiu-se de um modelo base com variáveis financeiras de controle e, por etapas, foram adicionadas novos determinantes e combinações dessas variáveis para montar modelos intermediários. Por fim, o modelo final contou com todas as variáveis explicativas mais relevantes. As variáveis de estudo podem ser classificadas em variáveis de estrutura de governança (DUA, GOV e COF), qualidade do conselho (QUA) e estrutura de propriedade (PRO1 e PRO2). Os modelos base utilizados foram: Daily e Dalton (1994a) e um próprio, desenvolvido para modelar melhor financial distress e sua relação com as variáveis de estrutura de governança. Nos diversos modelos testados foram encontradas relações significativas no percentual de conselheiros dependentes (GOV), percentual de conselheiros da elite educacional (QUA), percentual de ações discriminadas (PRO1) e percentual de ações de acionista estatal relevante (PRO2). Portanto, não se descartam as hipóteses de que mais conselheiros dependentes, menos conselheiros da elite educacional e estrutura de propriedade menos concentrada contribuem para uma situação de financial distress futura. Entretanto, as variáveis dummy de dualidade (DUA) e de conselho fiscal (COF) não apresentaram significância estatística.

Estruturas de governança corporativa e financial distress: há relação entre conselho de administração e empresas em financial distress?

Nesta dissertação foi analisada se há uma relação significante entre estruturas de governança (estrutura e composição de conselho) e financial distress. Este trabalho focou neste tema porque os estudos acadêmicos em governança corporativa e sua relação com financial distress ainda são pouco explorados. Além disso, o tema tem relevância no mundo corporativo, pois entender quais estruturas e composições de conselho seriam mais eficientes para evitar financial distress é interessante para diversos stakeholders, principalmente para os acionistas e os credores. Para verificar a existência dessa relação, foram utilizados dados de empresas brasileiras de capital aberto e foram desenvolvidos modelos logit de financial distress. Sendo a variável resposta financial distress, partiu-se de um modelo base com variáveis financeiras de controle e, por etapas, foram adicionadas novos determinantes e combinações dessas variáveis para montar modelos intermediários. Por fim, o modelo final contou com todas as variáveis explicativas mais relevantes. As variáveis de estudo podem ser classificadas em variáveis de estrutura de governança (DUA, GOV e COF), qualidade do conselho (QUA) e estrutura de propriedade (PRO1 e PRO2). Os modelos base utilizados foram: Daily e Dalton (1994a) e um próprio, desenvolvido para modelar melhor financial distress e sua relação com as variáveis de estrutura de governança. Nos diversos modelos testados foram encontradas relações significativas no percentual de conselheiros dependentes (GOV), percentual de conselheiros da elite educacional (QUA), percentual de ações discriminadas (PRO1) e percentual de ações de acionista estatal relevante (PRO2). Portanto, não se descartam as hipóteses de que mais conselheiros dependentes, menos conselheiros da elite educacional e estrutura de propriedade menos concentrada contribuem para uma situação de financial distress futura. Entretanto, as variáveis dummy de dualidade (DUA) e de conselho fiscal (COF) não apresentaram significância estatística

Estudo de estratégias de otimização para poços de petróleo com elevação por bombeio de cavidades progressivas

The progressing cavity pump artificial lift system, PCP, is a main lift system used in oil production industry. As this artificial lift application grows the knowledge of it s dynamics behavior, the application of automatic control and the developing of equipment selection design specialist systems are more useful. This work presents tools for dynamic analysis, control technics and a specialist system for selecting lift equipments for this artificial lift technology. The PCP artificial lift system consists of a progressing cavity pump installed downhole in the production tubing edge. The pump consists of two parts, a stator and a rotor, and is set in motion by the rotation of the rotor transmitted through a rod string installed in the tubing. The surface equipment generates and transmits the rotation to the rod string. First, is presented the developing of a complete mathematical dynamic model of PCP system. This model is simplified for use in several conditions, including steady state for sizing PCP equipments, like pump, rod string and drive head. This model is used to implement a computer simulator able to help in system analysis and to operates as a well with a controller and allows testing and developing of control algorithms. The next developing applies control technics to PCP system to optimize pumping velocity to achieve productivity and durability of downhole components. The mathematical model is linearized to apply conventional control technics including observability and controllability of the system and develop design rules for PI controller. Stability conditions are stated for operation point of the system. A fuzzy rule-based control system are developed from a PI controller using a inference machine based on Mandami operators. The fuzzy logic is applied to develop a specialist system that selects PCP equipments too. The developed technics to simulate and the linearized model was used in an actual well where a control system is installed. This control system consists of a pump intake pressure sensor, an industrial controller and a variable speed drive. The PI control was applied and fuzzy controller was applied to optimize simulated and actual well operation and the results was compared. The simulated and actual open loop response was compared to validate simulation. A case study was accomplished to validate equipment selection specialist system

Control automatico de un sistema electrico mediante logica difusa

Substitution of fuzzy logic control in an electrical system normally controlled by proportional-integral frequency was studied and analyzed. A linear model of an electrical system, the concepts which govern the theory of fuzzy logic, and the application of this theory to systems control, are briefly presented. The methodology of fuzzy logic was then applied to develop a model for an electrical energy system. The results of the simulation demonstrated that fuzzy logic control eliminated the area frequency error and permitted that only the area experiencing an increase in charge responds to this variation. Based on the results, it is concluded that control based on fuzzy logic is simple, is easy to maintain, is of low cost, and can be used to substitute traditional velocity controllers.

Influência do formato do rebordo alveolar na distribuição interna das tensões em prótese parcial removível dento-implantossuportada

Pós-graduação em Odontologia - FOA

Implementação de um sistema para geração e avaliação de movimentos em pacientes hemiplégicos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Coordenação de relés de sobrecorrente e distância para sistemas de transmissão

Pós-graduação em Engenharia Elétrica - FEIS

Protein folding kinetics exhibit an Arrhenius temperature dependence when corrected for the temperature dependence of protein stability

The anomalous temperature dependence of protein folding has received considerable attention. Here we show that the temperature dependence of the folding of protein L becomes extremely simple when the effects of temperature on protein stability are corrected for; the logarithm of the folding rate is a linear function of 1/T on constant stability contours in the temperature–denaturant plane. This convincingly demonstrates that the anomalous temperature dependence of folding derives from the temperature dependence of the interactions that stabilize proteins, rather than from the super Arrhenius temperature dependence predicted for the configurational diffusion constant on a rough energy landscape. However, because of the limited temperature range accessible to experiment, the results do not rule out models with higher order temperature dependences. The significance of the slope of the stability-corrected Arrhenius plots is discussed.

Progression of human breast cancers to the metastatic state is linked to hydroxyl radical-induced DNA damage.

Hydroxyl radical damage in metastatic tumor DNA was elucidated in women with breast cancer, and a comparison was made with nonmetastatic tumor DNA. The damage was identified by using statistical models of modified base and Fourier transform-infrared spectral data. The modified base models revealed a greater than 2-fold increase in hydroxyl radical damage in the metastatic tumor DNA compared with the nonmetastatic tumor DNA. The metastatic tumor DNA also exhibited substantially greater base diversity than the nonmetastatic DNA, and a progression of radical-induced base damage was found to be associated with the growth of metastatic tumors. A three-dimensional plot of principal components from factor analysis, derived from infrared spectral data, also showed that the metastatic tumor DNA was substantially more diverse than the tightly grouped nonmetastatic tumor DNA. These cohesive, independently derived findings suggest that the hydroxyl radical generates DNA phenotypes with various metastatic potentials that likely contribute to the diverse physiological properties and heterogeneity characteristic of metastatic cell populations.

Data mining and simulation: a grey relationship demonstration

Fuzzy data has grown to be an important factor in data mining. Whenever uncertainty exists, simulation can be used as a model. Simulation is very flexible, although it can involve significant levels of computation. This article discusses fuzzy decision-making using the grey related analysis method. Fuzzy models are expected to better reflect decision-making uncertainty, at some cost in accuracy relative to crisp models. Monte Carlo simulation is used to incorporate experimental levels of uncertainty into the data and to measure the impact of fuzzy decision tree models using categorical data. Results are compared with decision tree models based on crisp continuous data.

Combined Machine Learning Techniques for Decision Making Support in Medicine

Computational intelligent support for decision making is becoming increasingly popular and essential among medical professionals. Also, with the modern medical devices being capable to communicate with ICT, created models can easily find practical translation into software. Machine learning solutions for medicine range from the robust but opaque paradigms of support vector machines and neural networks to the also performant, yet more comprehensible, decision trees and rule-based models. So how can such different techniques be combined such that the professional obtains the whole spectrum of their particular advantages? The presented approaches have been conceived for various medical problems, while permanently bearing in mind the balance between good accuracy and understandable interpretation of the decision in order to truly establish a trustworthy ‘artificial’ second opinion for the medical expert.

A New Gradient Descent Approach for Local Learning of Fuzzy Neural Models

The majority of reported learning methods for Takagi-Sugeno-Kang fuzzy neural models to date mainly focus on the improvement of their accuracy. However, one of the key design requirements in building an interpretable fuzzy model is that each obtained rule consequent must match well with the system local behaviour when all the rules are aggregated to produce the overall system output. This is one of the distinctive characteristics from black-box models such as neural networks. Therefore, how to find a desirable set of fuzzy partitions and, hence, to identify the corresponding consequent models which can be directly explained in terms of system behaviour presents a critical step in fuzzy neural modelling. In this paper, a new learning approach considering both nonlinear parameters in the rule premises and linear parameters in the rule consequents is proposed. Unlike the conventional two-stage optimization procedure widely practised in the field where the two sets of parameters are optimized separately, the consequent parameters are transformed into a dependent set on the premise parameters, thereby enabling the introduction of a new integrated gradient descent learning approach. A new Jacobian matrix is thus proposed and efficiently computed to achieve a more accurate approximation of the cost function by using the second-order Levenberg-Marquardt optimization method. Several other interpretability issues about the fuzzy neural model are also discussed and integrated into this new learning approach. Numerical examples are presented to illustrate the resultant structure of the fuzzy neural models and the effectiveness of the proposed new algorithm, and compared with the results from some well-known methods.

Design of neuro-fuzzy models by evolutionary and gradient-based algorithms

All systems found in nature exhibit, with different degrees, a nonlinear behavior. To emulate this behavior, classical systems identification techniques use, typically, linear models, for mathematical simplicity. Models inspired by biological principles (artificial neural networks) and linguistically motivated (fuzzy systems), due to their universal approximation property, are becoming alternatives to classical mathematical models. In systems identification, the design of this type of models is an iterative process, requiring, among other steps, the need to identify the model structure, as well as the estimation of the model parameters. This thesis addresses the applicability of gradient-basis algorithms for the parameter estimation phase, and the use of evolutionary algorithms for model structure selection, for the design of neuro-fuzzy systems, i.e., models that offer the transparency property found in fuzzy systems, but use, for their design, algorithms introduced in the context of neural networks. A new methodology, based on the minimization of the integral of the error, and exploiting the parameter separability property typically found in neuro-fuzzy systems, is proposed for parameter estimation. A recent evolutionary technique (bacterial algorithms), based on the natural phenomenon of microbial evolution, is combined with genetic programming, and the resulting algorithm, bacterial programming, advocated for structure determination. Different versions of this evolutionary technique are combined with gradient-based algorithms, solving problems found in fuzzy and neuro-fuzzy design, namely incorporation of a-priori knowledge, gradient algorithms initialization and model complexity reduction.