by Chow Kan Shing. / Includes questionnaire in Chinese. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1993. / Includes bibliographical references (leaves 100-103). / Chapter Chapter 1. --- Introduction --- p.1 / Chapter Chapter 2. --- A Brief Review on Fuzzy Set Theory --- p.3 / Chapter 2.1. --- The Concept of Fuzzy Set Theory --- p.3 / Chapter 2.2. --- Fundamental Operations on Fuzzy Sets --- p.4 / Chapter 2.3. --- Two Approaches to Investigate Fuzzy Set Theory --- p.6 / Chapter Chapter 3. --- The Interpretation of the Membership Function --- p.7 / Chapter 3.1. --- Review and Comparison of the Interpretation of the Membership Values --- p.7 / Chapter 3.1.1. --- Interpretation in terms of Betting / Chapter 3.1.2. --- Interpretation in terms of Payoff Function / Chapter 3.1.3. --- Interpretation in terms of Amount of Relevant Attribute / Chapter 3.1.4. --- Interpretation in terms of the TEE Model / Chapter 3.1.5. --- Interpretation in terms of a Measurement Model / Chapter 3.1.6. --- Interpretation in terms of Prototype Theory / Chapter 3.2. --- Discussion about Membership Function --- p.29 / Chapter Chapter 4. --- Estimation of the Membership Function --- p.33 / Chapter 4.1. --- The Data Collection Methods for the Estimation of the Membership Function --- p.34 / Chapter 4.1.1. --- Direct Rating / Chapter 4.1.2. --- Polling / Chapter 4.1.3. --- Set-valued Statistics / Chapter 4.1.4. --- Reverse Rating / Chapter 4.2. --- Estimation Procedures for the Membership Function and their Characteristics --- p.36 / Chapter 4.2.1. --- Non-parametric Estimation Procedures / Chapter 4.2.2. --- The Characteristics of the Non-parametric Estimation Procedures / Chapter 4.2.3. --- Parametric Estimation Procedures / Chapter 4.3. --- Connections between the Four Data Collection Methods --- p.58 / Chapter 4.3.1. --- Connection between Direct Rating and Polling / Chapter 4.3.2. --- Connection between Polling and Reverse Rating / Chapter 4.3.3. --- Connection between Reverse Rating and Set-valued Statistics / Chapter 4.4. --- Other Estimation Procedures --- p.71 / Chapter 4.4.1. --- Procedure based on Saaty's Matrix / Chapter 4.4.2. --- Procedure based on Mabuchi's Interpretation of the Membership Function / Chapter 4.5. --- The Survey --- p.77 / Chapter 4.5.1. --- Introduction of the Survey / Chapter 4.5.2. --- The Result of the Survey / Chapter 4.5.3. --- An Approach to reduce the 'bias' in Polling / Chapter 4.5.4. --- Advice to Researchers / Chapter Chapter 5. --- Discussion --- p.97 / References --- p.100 / Appendix: Questionnaire

Orientadores: Pedro Aladar Tonelli, Rodney Carlos Bassanezi / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Matematica, Estatistica e Computação Cientifica / Made available in DSpace on 2018-09-10T15:15:48Z (GMT). No. of bitstreams: 1 Barros_LaecioCarvalhode_D.pdf: 1779072 bytes, checksum: 765be2e1f0125564052242265b762ff9 (MD5) Previous issue date: 1997 / Resumo: Não informado / Abstract: Not informed / Doutorado / Doutor em Matemática Aplicada

Thesis (Ph. D.)--University of Alabama at Birmingham, 2007. / Print out. Additional advisors: James J. Buckley, Jeffrey G. Gray, Robert M. Hyatt, Randy K. Smith. Includes bibliographical references (leaves 185-207). Also available via the World Wide Web.

This research extends fuzzy methods to consider the fuzzy validation of fuzzy land cover data at the sub-pixel level. The study analyses the relationships between fuzzy memberships generated by field survey and those generated from the classification of remotely sensed data. In so doing it examines the variations in the relationship between observed and predicted fuzzy land cover classes. This research applies three land cover classification techniques: Fuzzy sets, Fuzzy c-means and Boolean classification, and develops three models to determine fuzzy land cover change. The first model is dependent on fuzzy object change. The second model depends on the sub-pixel change through a fuzzy change matrix, for both fuzzy sets and fuzzy c-means, to compute the fuzzy change, fuzzy loss and fuzzy gain. The third model is a Boolean change model which evaluates change on a pixel-by-pixel basis. The results show that using a fuzzy change analysis presents a subtle way of mapping a heterogeneous area with common mixed pixels. Furthermore, the results show that the fuzzy change matrix gives more detail and information about land cover change and is more appropriate than fuzzy object change because it deals with sub-pixel change. Finally the research has found that a fuzzy error matrix is more suitable than an error matrix for soft classification validation because it can compare the membership from the field with the classified image. From this research there arise some important points: • Fuzzy methodologies have the ability to define the uncertainties associated with describing the phenomenon itself and the ability to take into consideration the effect of mixed pixels. • This research compared fuzzy sets and fuzzy c-means, and found the fuzzy set is more suit-able than fuzzy c-means, because the latter suffers from some disadvantages, chiefly that the sum of membership values of a data point in all the clusters must be one, so the algorithm has difficulty in handling outlying points. • This research validates fuzzy classifications by determining the fuzzy memberships in the field and comparing them with the memberships derived from the classified image.

Currently, capacity planning is receiving more emphasis in management of operations in Industrial Engineering because insufficient capacity may lead to deteriorating delivery performance and high work-in-process inventories. On the other hand excess capacity may lead to wastage of resources. Even the most modern and sophisticated capacity planning systems may face a great deal of uncertainty, imprecision and vagueness due to uncertain market demand, set up resources, capacity constraints, pessimistic time standards, and subjective beliefs of managers etc., leading to inferior planning decisions. Under such circumstances fuzzy models which explicitly consider these uncertainties, generate more robust, flexible and efficient planning. The traditional fuzzy logic-based models though are capable of dealing with some complex capacity-planning systems where various uncertain parameters and vagueness are involved, yet they use complex membership functions to calculate the degree of truth that involve complicated, time consuming and tedious mathematical operations. In this thesis, the solution techniques and methods developed are based on possibility theory. These techniques not only eliminate the need of calculation of complex membership functions but also yield crisp answers to fuzzy problems in capacity planning.

Currently, capacity planning is receiving more emphasis in management of operations in Industrial Engineering because insufficient capacity may lead to deteriorating delivery performance and high work-in-process inventories. On the other hand excess capacity may lead to wastage of resources. Even the most modern and sophisticated capacity planning systems may face a great deal of uncertainty, imprecision and vagueness due to uncertain market demand, set up resources, capacity constraints, pessimistic time standards, and subjective beliefs of managers etc., leading to inferior planning decisions. Under such circumstances fuzzy models which explicitly consider these uncertainties, generate more robust, flexible and efficient planning. The traditional fuzzy logic-based models though are capable of dealing with some complex capacity-planning systems where various uncertain parameters and vagueness are involved, yet they use complex membership functions to calculate the degree of truth that involve complicated, time consuming and tedious mathematical operations. In this thesis, the solution techniques and methods developed are based on possibility theory. These techniques not only eliminate the need of calculation of complex membership functions but also yield crisp answers to fuzzy problems in capacity planning.

Thesis (M.S.)--Youngstown State University, 1996. / Includes bibliographical references (leaves 57-59)

Thesis (Ph. D.)--University of Washington, 2004. / Vita. Includes bibliographical references (p. 136-138).

Orientador: João de Deus Mendes da Silva / Dissertação (mestrado profissional) - Universidade Estadual de Campinas, Instituto de Matematica, Estatistica e Computação Cientifica / Made available in DSpace on 2018-08-15T01:57:01Z (GMT). No. of bitstreams: 1 Barros_AntonioMagno_M.pdf: 1363614 bytes, checksum: d02b8dd0f8f8eafcf19aab637ef36f26 (MD5) Previous issue date: 2009 / Resumo: Os fenômenos epidemiológicos apresentam vários tipos de subjetividades, nas quais, em muitas ocasiões, são tratadas de maneira eficiente pelos modelos clássicos. Entretanto, a lógica fuzzy se apresenta de maneira adequada para tratar tais subjetividades. Neste trabalho, realizamos um estudo sobre os modelos epidemiológicos do tipo SI, SIS e SIR. Em seguida apresentamos os principais conceitos da teoria dos conjuntos fuzzy, controladores fuzzy e sistemas dinâmicos p-fuzzy. Fazemos, também um estudo dos modelos epidemiológicos fuzzy onde utilizamos o valor esperado fuzzy como defuzificador. Por fim, propomos uma comparação entre os modelos clássicos, p-fuzzy e valor esperado fuzzy. / Abstract: The epidemiological phenomena have several types of subjectivities, in which, on many occasions, are handled efficiently by classical models. However, fuzzy logic is presented properly to treat such subjectivities. We carried out a study on the epidemiological models of type SI, SIS and SIR. The following are the main concepts of the theory of fuzzy sets, fuzzy controllers and p-fuzzy dynamic systems. We are also a study of epidemiological models where we use the fuzzy expected value as fuzzy defuzificador. Finally, we propose a comparison between the classical models, p-fuzzy and fuzzy expected value. / Mestrado / Biomatematica / Mestre em Matemática

[pt] Um controlador Heurístico Adaptativo baseia-se num conjunto de regras lingüísticas para conduzir um processo com modelo impreciso ou complexo ao estado desejado. O comportamento do processo deve respeitar os requisitos de performance predefinidos. Para satisfazer estes objetivos, a estrutura interna do controle sofre mudanças para adequá- la as condições vigentes no processo. Os métodos de adaptação abordados consideram a modificação de uma estrutura matricial interpretada como as correções incrementais, compatíveis com os ajustes a serem efetuados sobre o processo, ou como regras, constituídas por variáveis nebulosas, que requerem manipulações adicionais para produzir a saída do controlador. Em qualquer dos casos, a adaptação é realizada a partir de uma Tabela de Índices de Performance. Para facilitar a sua obtenção é implementado um procedimento, que fornece a representação matricial das regras lingüísticas, concatenadas na forma de um Algoritmo Lingüístico de Controle. O comportamento dinâmico do Sistema, composto pelos Controladores Heurísticos e por processos com modelos distintos, é considerado para Tabelas de índices de Performance com várias dimensões. As regras lingüísticas, correlacionadas com estas tabelas, foram elaboradas com diversas classes de atributos. As simulações realizadas concentram-se sobre os parâmetros dos controladores, que influenciam significativa- Os estudos abordam também o comportamento da estrutura interna destes controladores e o seu desempenho em termos da velocidade de atuação sobre o processo. / [en] A heuristic Controller uses a set of linguistic rules, which are derived from expertise or human operators´ skills, in order to achieve control of processes that have inaccurate or complex models. An adaptative Heuristic Controller adjusts the set of rules in an automatic and continuous way, aiming to achieve prescribed objectives indicated by a performance measure. The adaptative procedures modify a matrix, the elements of which are either incremental corrections or numeric rules associated with fuzzy variables. In both cases a Performance Index Table and a learning method are employed to correct that matrix. The Performance Table is a matrix calculated from a set of linguistic rules. The controllers are implemented with different Performance Tables, considering various sets of linguistic values and quantization levels. The dynamic behaviour of overdamped and underdamped processes is investigated. The performance of simulated systems is analyzed with respect to relevant parameters that affect their behaviour.