Enthalpy-based decision trees for comfort assessment of light layers in a tropical environment

• Autor(a) principal: Harada, Érik dos Santos
• Data de Publicação: 2021
• Outros Autores: Montanhani , Maria Elena Silva, Bueno, Leda Gobbo de Freitas, Mollo Neto, Mario, Souza, Silvia Regina Lucas de, Fonseca, Ricardo da
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Research, Society and Development
• Texto Completo: https://rsdjournal.org/index.php/rsd/article/view/13354
• topic: Climate extremes; Data mining; Layer poultry.; Extremos Climáticos; Minería de datos; Aves de Corral ponedoras.; Extremos climáticos; Mineração de dados; Avicultura de postura.

With the advancement of technology, it was possible to store a large amount of data at a lower price, making it possible to investigate the climatic factors that affect animal production. The western region of the state of São Paulo is a region prone to extremes of temperature, which is a worrying factor for egg producers. This study aimed to find out if it is possible to generate decision trees from enthalpy data using Data Mining, and if these trees are suitable to be inserted in a weather forecast system. For this study, a database of bioclimatic variables from three commercial layer aviaries located in the city of Bastos-SP, Brazil, in the year 2013 was used. After organizing and classifying the data in a spreadsheet, and processed by Weka software with the J48 algorithm (C4.5) for data mining, this technique applied to this database allowed the generation of decision trees with approximately 98% and 0.96 for Kappa index, respectively. Thus, the decision trees generated in this study are accurate enough to be used in a future warning system against climatic adversities for commercial layers in tropical climates.