Utilização da modelagem matemática (redes neurais artificiais) na classificação de autotetraploides de bananeira (Musa acuminata Colla)

Oliveira, Ana Catarina Lima de; Pasqual, Moacir; Pio, Leila Aparecida Salles; Lacerda, Wilian Soares; Oliveira e Silva, Sebastião de

Repositório Institucional

Use este identificador para citar ou linkar para este item: https://repositorio.ifs.edu.br/biblioteca/handle/123456789/621

Registro completo de metadados

Campo DC	Valor	Idioma
dc.date.accessioned	2018-07-11T12:48:52Z	-
dc.date.available	2018-07-11T12:48:52Z	-
dc.date.issued	2013-05	-
dc.identifier.citation	OLIVEIRA, Ana Catarina Lima de [et al]. Utilização da modelagem matemática (redes neurais artificiais) na classificação de autotetraploides de bananeira (Musa acuminata Colla). Bioscience Journal (Online), v. 29, p. 617-622, 2013.	pt_BR
dc.identifier.issn	1981-3163	-
dc.identifier.uri	https://repositorio.ifs.edu.br/biblioteca/handle/123456789/621	-
dc.description.abstract	Objetivou-se desenvolver uma metodologia para possibilitar a classificação de plantas de bananeira submetidas à indução de duplicação cromossômica utilizando Redes Neurais Artificiais (RNA). Os dados utilizados neste trabalho foram retirados de uma tese já apresentada, cujo autor estudou a correlação entre a massa fresca de discos foliares e o conteúdo de DNA. A RNA foi implementada com a função de classificação. A taxa de aprendizado e o termo momentum adotados foram respectivamente iguais a 0,01 e 0,2, o número de épocas de treinamento foi 1000. Esses valores foram determinados por meio de tentativa e erro. Para o treinamento, 90% das plantas foram utilizadas e, para validação, 10% do total de 114 autotetraploides produzidos artificialmente por meio de exposição ao antimitódico colchicina. A RNA classificou corretamente 10 das 11 amostras utilizadas para validação. A estatística Kappa foi de 63,33%, o que indica que a RNA pode ainda ser melhorada. A rede neural artificial do tipo Multi Layer Perceptron implementada é eficaz na pré-seleção de poliploides desejáveis de bananeira Tong Dok Mak.	pt_BR
dc.language.iso	pt_BR	pt_BR
dc.subject	Ciências agrárias	pt_BR
dc.subject	Agronomia	pt_BR
dc.subject	Musa acuminata	pt_BR
dc.subject	Redes neurais artificiais	pt_BR
dc.subject	Autotetraploides	pt_BR
dc.subject	Modelagem matemática	pt_BR
dc.subject	Bananeira	pt_BR
dc.subject	Artificial neural net works	en
dc.subject	Mathematical modeling	en
dc.subject	Agronomy	en
dc.subject	Agrarian science	en
dc.title	Utilização da modelagem matemática (redes neurais artificiais) na classificação de autotetraploides de bananeira (Musa acuminata Colla)	pt_BR
dc.title.alternative	Use of mathematical modeling (artificial neural networks) in classification of banana autotetraploid (Musa acuminata COLLA)	en
dc.type	Artigo	pt_BR
dcterms.title	Bioscience journal	-
dc.contributor.author	Oliveira, Ana Catarina Lima de	-
dc.contributor.author	Pasqual, Moacir	-
dc.contributor.author	Pio, Leila Aparecida Salles	-
dc.contributor.author	Lacerda, Wilian Soares	-
dc.contributor.author	Oliveira e Silva, Sebastião de	-
dc.contributor	Machado, Ana Catarina Lima de Oliveira	-
dc.description.abstract2	The objective was to develop a methodology to enable the classification of banana plants submitted to the induction of chromosome duplication using Artificial Neural Networks (RNA). The data used in this work were taken from a thesis already presented, whose author studied the correlation between the fresh mass of leaf discs and the DNA content. RNA was implemented with the classification function. The learning rate and the term momentum were respectively 0.01 and 0.2, the number of training times was 1000. These values were determined by trial and error. For the training, 90% of the plants were used and, for validation, 10% of the total of 114 autotetraploids produced artificially through exposure to the antimitodic colchicine. RNA correctly classified 10 of the 11 samples used for validation. The Kappa statistic was 63.33%, which indicates that the RNA can still be improved. The artificial multi-layered Perceptron neural network implemented is effective in pre-selecting the desirable polyploid polyploids of the Tong Dok Mak banana tree.	en
dc.description.abstract2	UHT milk is defined as the homogenised milk which has been subjected for 2 to 4 seconds to a temperature between 130 ° C and 150 ° C, before a continuous flow heat process, cooled to a temperature below 32 ° C and packed under sterile conditions in aseptic packages and hermetically sealed by specialized machines in this type of production. Due to the high consumption of UHT milk in the world, it is important to evaluate its quality so that the product reaches the consumer safe and within the standards recommended by the current legislation. The objective of this study was to evaluate the physical-chemical and microbiological characteristics of three brands of UHT milk sold in the city of Nossa Senhora da Glória - SE. The experiments were carried out at the Multifunctional Dairy Laboratory of the Federal Institute of Sergipe Campus Glória. Three brands of integral UHT milk marketed in the municipality of Nossa Senhora da Glória - SE were used. The samples were purchased at the local commerce of the municipality and transported to the IFS Campus Glória Laboratory. The samples were duly coded as "A", "B" and "C" for the purpose of maintaining the integrity of the companies. In view of the results obtained in the physicochemical analyzes, it was concluded that of the three brands analyzed, brand "A" presented lower values of pH and index cryoscope than is recommended by current legislation. The "B" and "C" marks presented results within the limits established by the legislation. In relation to the microbiological analyzes, all the analyzed brands met the standards established by Ordinance No. 370 of 1997. It is suggested with this work the importance of standardizing the manufacturing process in order to meet the parameters recommended by the legislation. In addition, the adoption of BPA, GMP and HACCP throughout the milk production chain.	en
Aparece nas coleções:	Artigo, Resumo científico e Comunicação em eventos - Agroecologia

Arquivos associados a este item:

Arquivo	Descrição	Tamanho	Formato
Utilização da modelagem matemática (redes neurais artificiais) na classificação de autotetraploides de bananeira (Musa acuminata Colla).pdf		93,4 kB	Adobe PDF	Visualizar/Abrir

Mostrar registro simples do item Visualizar estatísticas

Ferramentas do administrador