Título: Redes Neurais Artificiais: Aplicação na Predição da Percentagem de Carne na Carcaça Suína Resfriada
Autores: Dartora, Gery A.; Borges, Paulo S. S.; Tápia, Milena; Guidonni, Antônio
Resumo: This paper intends to discuss and evaluate the ability of the Artificial Neural Networks (ANNs) in accomplishing accurate percentage of meet in swine carcass forecast. This variable is of great importance in the bonus rate determination used in the carcass tipification processs. The study is justified due to the importance of the problem to the development of Brazilian swine culture and the low quality prediction obtained by the actual model (Multiple Regression – Least Squares Estimation). The evaluation of the model is did comparing the results obtained by the ANN with those reached by the application of the Least Squares Method. The experiments indicate that the ANN, in spite of been based on data – limiting factor in this work – presents the bigger predictive capacity.
Código DOI: 10.21528/CBRN2003-047
Artigo em PDF: 6CBRN_047.PDF
Arquivo BibTex: 6CBRN_047.bib