Comparing Non-Linear Models to Describe the Growth Performance in Large White Yorkshire Pigs
Archana K. Selvam S. Serma SaravanaPandian A. Sivakumar T. Balasubramanyam D. Mohamed Safiullah A
Vol 7(11), 148-152
Exotic pigs are potential animal protein suppliers in developing country like India. An attempt was made to apply different nonlinear growth models to predict the best growth model for Large White Yorkshire pigs and to examine the difference in the growth pattern of male and female piglets. The data pertaining to achieve the objectives were collected between 2011 to 2016 at PGRIAS, Kattupakkam, on bodyweights at different ages of pigs with the sample size of 1226 Male and 1120 Female Large White Yorkshire pigs were included in the study. The data on body weights were plotted against the age for both males and females separately, and the shape of the curve was examined. Based on the shape of the curves, the following models were fitted to the data viz. Brody model, Logistic model, Gompertz model, Von Bertalanffy model and Richard model. For non linear models the highest R square value, Root Mean Square Error (RMSE), Mean Absolute Error (MAE) and Mean Absolute Predicted Error (MAPE) criteria were used to determine the best fitted model. The analysis of data revealed that in male pigs the value of highest R2 (0.999) and least MAE (0.597), RMSE (1.755), MAPE (29.85) were observed in Gompertz model followed by Logistic model. In case of female pigs the highest R2 (0.998) and least MAE (0.654), RMSE (1.130), MAPE (29.03) were seen for Gompertz model. Thus, it could be concluded that as the best fit for assessing growth of male and female Large White Yorkshire pigs.
Keywords : Growth Performance Large White Yorkshire Non –Linear Model
Read : 424
Downloads : 119