Gpfis: Um Sistema Fuzzy-Genético Genérico Baseado Em Programação Genética

KOSHIYAMA, ADRIANO SOARES

doi:10.17771/pucrio.acad.26560

Cited by 7 publications

(17 citation statements)

References 125 publications

(272 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Explora-se, nesta dissertação, o GPFIS-Forecast: versão do GPFIS Koshiyama et al (2013Koshiyama et al ( , 2014; Koshiyama (2014); Koshiyama et al (2015) para previsão de série temporais univariada. Cada etapa desse modeloé repleta de possibilidades de diferentes configurações.…”

Section: Motivaçãounclassified

“…Os Preditores Fuzzy (PF) para séries temporais são muito utilizados atualmente (Wang e Mendel, 1992;Kim e Kim, 1997;Mangai et al, 2014;Stepnicka et al, 2014;Lu et al, 2014;Bergmeir et al, 2012;Hadavandi et al, 2011;Koshiyama, 2014), fornecendo resultados que podem ser explicados, favorecendo a análise e viabilizando planejamentos estratégicos.…”

Section: Previsor Fuzzy Para Séries Temporaisunclassified

“…Uma das variantes do GPFIS (Koshiyama, 2014) contempla aárea de previsão de séries temporais univariadas; o modelo para talé denominado GPFIS-Forecast. Este capítulo se inicia com a descrição básica da estrutura deste modelo.…”

Section: Modelo Gpfis -Forecastunclassified

“…Tomando por base a configuração do melhor experimento de Koshiyama (2014), cujos parâmetros são apresentados na Tabela 4.4, avaliou-se o desempenho do modelo para 500, 250, 100, 50 gerações da MHG na previsão das 11últimas séries da NN3.…”

Section: Experimentosunclassified

See 3 more Smart Citations

Gpfis-Forecast: Um Sistema Fuzzy-Genético Baseado Em Programação Genética Multigênica Para Problemas De Previsão Univariada

FERREIRA¹

View full text Add to dashboard Cite

Ferreira, Marco Antônio da Cunha; Tanscheit, Ricardo (Advisor); Vellasco, Marley Maria Bernardes Rebuzzi Vellasco (Co-Advisor). GPFIS-Forecast: A Genetic-Fuzzy system based on genetic programming for forecast problems. Rio de Janeiro, 2015. 55p. MSc. Dissertation -Departamento de Engenharia Elétrica, Pontifícia Universidade Católica do Rio de Janeiro.Forecasting methods are very important for the development of various activities in everyday society. Several statistical models have been developed, but many assumptions must be made in order to obtain an acceptable response. Nonstatistical models for time series forecasting such as those involving systems Fuzzy Inference Systems (FIS) provide a description of the process through linguistic rules. This dissertation delves into GPFISForecast: a version of GPFIS -Fuzzy Inference System based on Multigene Genetic Programming -for univariate time series forecasting. This model consists of four basic stages: Fuzzification, Inference, Defuzzification and Evaluation. In each of these steps, different configurations will have distinct impacts on the results. This work proposes the improvement of GPFIS-Forecast along two main lines (i) increase the amount of possible configurations and assess their contribution to a better forecasting accuracy and (ii) add further information to the interpretation of results, keeping in mind both accuracy and interpretability. The case studies show that in the case of time series with small tendency, GPFIS-Forecast provides a good accuracy; when tendency is larger and pre-processing becomes necessary, interpretability is affected. The Fuzzy Forecasting Limits introduced here add more information to the result, pointing to possible adjustments to rule bases of models with greater granularity.

show abstract

Section: Motivaçãounclassified

Section: Previsor Fuzzy Para Séries Temporaisunclassified

Section: Modelo Gpfis -Forecastunclassified

Section: Experimentosunclassified

See 2 more Smart Citations

Gpfis-Forecast: Um Sistema Fuzzy-Genético Baseado Em Programação Genética Multigênica Para Problemas De Previsão Univariada

FERREIRA¹

View full text Add to dashboard Cite

show abstract

“…-O modelo GPFIS [18] (Genetic Programing Fuzzy Inference System), inspirado em [33,36,37], surge como uma arquitetura genérica do tipo Pittsburgh que pode atuar em diferentes aplicações, e não somente em classificação. A definição da classe que melhor se ajusta ao antecedente da regra é realizada por uma heurística, prescindido de uma busca via PGMG.…”

Section: Abordagens Para Base De Regrasunclassified

Síntese Automática De Sistemas De Inferência Fuzzy Para Classificação

MERINO¹

View full text Add to dashboard Cite

Automatic Synthesis of Fuzzy Inference Systems for Classification

Paredes

Tanscheit

Vellasco

et al. 2016

Information Processing and Management of Uncertainty in Knowledge-Based Systems

View full text Add to dashboard Cite

Nowadays, much of the accumulated knowledge is stored as data. In many classification problems the relationship between a set of variables (attributes) and a target variable of interest must be learned. Among the tools capable of modeling real systems, Fuzzy Inference Systems are considered excellent with respect to the knowledge representation in a comprehensible way, as they are based on inference rules. This is relevant in applications where a "black box" model does not suffice. This model may attain good accuracy, but does not explain how results are obtained. This dissertation presents the development of a Fuzzy Inference System in an automatic manner, where the rule base should favour linguistic interpretability and at the same time provide good accuracy. In this sense, this work proposes the AutoFIS-Class model, an automatic method for generating Fuzzy Inference Systems for classification problems. Its main features are: (i) generation of premises to ensure minimum, quality criteria, (ii) association of each rule premise to the most compatible consequent term; and (iii) aggregation of rules for each class through operator that weigh the relevance of each rule. The proposed model was evaluated for 45 datasets and their results were compared to existing models based on Evolutionary Algorithms. Results show that the proposed Fuzzy Inference System is competitive, presenting good accuracy with a low number of rules.

show abstract

Gpfis: Um Sistema Fuzzy-Genético Genérico Baseado Em Programação Genética

Cited by 7 publications

References 125 publications

Gpfis-Forecast: Um Sistema Fuzzy-Genético Baseado Em Programação Genética Multigênica Para Problemas De Previsão Univariada

Gpfis-Forecast: Um Sistema Fuzzy-Genético Baseado Em Programação Genética Multigênica Para Problemas De Previsão Univariada

Síntese Automática De Sistemas De Inferência Fuzzy Para Classificação

Automatic Synthesis of Fuzzy Inference Systems for Classification

Contact Info

Product

Resources

About