Desenvolvimento de aplicações em medicina e agronomia utilizando lógica fuzzy e neuro fuzzy

• Ano de defesa: 2014
• Autor(a) principal: Silva, Aldo Antonio Vieira da [UNESP]
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual Paulista (Unesp)
• Palavras-chave em Português: Lógica difusa; Agronomia; Medicina; Adubação; Fuzzy logic
• Link de acesso: http://hdl.handle.net/11449/110517

This work proposes two new application methods: one in the area of biomedical engineering in the diagnosis of inguinal hernias using fuzzy logic and another in the area of agriculture to estimate the wheat productivity using an adaptive neuro fuzzy inference system. The first was an application developed for mobile devices, smartphones and tablets, to assist decision making in the diagnosis of patients with suspected inguinal hernia. It was used the Java language together with the fuzzy logic library, denominated jfuzzylogic and the Android operating system for the application development. To validate the application it was used data obtained via questionnaire, involving 30 patients interviewed in medical consultation. As a result, it was observed that the diagnosis made by the medical team and diagnosis with the aid of the mobile application, were equivalent in cases of affected patients with hernia in the inguinal region. This software is available free of charge via the web, for professionals in the health field. In the second application method, it was investigated the ability to develop an adaptive neuro fuzzy inference system for estimating the productivity of wheat (Triticum aestivum) in relation to the nitrogen fertilization, based on experimental data of wheat cultivars during two years, in Selvíria-MS. Through the data input and output, the system of adaptive neuro fuzzy inference learns and subsequently can estimate a new value of wheat production based on different doses of nitrogen. The results showed that the neuro fuzzy system is feasible to develop a prediction model to estimate the productivity of wheat in relation to nitrogen rates. The RMSE (Root Mean Square Error) error of the estimated wheat productivity using the neuro fuzzy system was smaller than that obtained with the quadratic regression method, that is usually used in this kind of estimated, and also the relation between ...