Jonatas Cattelam, Marcos M. do Vale, Patrícia M. Martini, Rangel F. Pacheco, Andrei R. Mayer and Paulo S. Pacheco
The aim of this research was to characterize studies on production of super precocious and precocious cattle by meta-analytical approach associated with data mining, using articles published between 2002 and 2012. The selected research was intended to address the production of male bovines, receiving the same level of concentrate during all experimental period, receiving as roughage, exclusively corn silage. Data were classified as super precocious for cattle until 18 months of age at the end of the experiment; and precocious, for bovines with aged between 19 and 30 months to the finishing of the experiment. The classification accuracy was 89% with accuracies from 89% for super precocious and 92% for precocious. The built model showed the number of days confined as root node, which is the primary node in the decision tree. System production of super precocious steers require greater confinement period with high share of food with low fiber and high energy and high energy level of the diet. The exploitation of results through the decision tree model features easy viewing and interpretation, with good implementation capacity in animal production.
Confinement days, Data mining, Decision tree, Meta-analysis