Comparative Analysis of Artificial Intelligence and Nonlinear Models for Broiler Growth Curve
Uluslararası Tarım ve Yaban Hayatı Bilimleri Dergisi, cilt.7, sa.3, ss.515-523, 2021 (TRDizin)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 7 Sayı: 3
- Basım Tarihi: 2021
- Doi Numarası: 10.24180/ijaws.990297
- Dergi Adı: Uluslararası Tarım ve Yaban Hayatı Bilimleri Dergisi
- Derginin Tarandığı İndeksler: TR DİZİN (ULAKBİM)
- Sayfa Sayıları: ss.515-523
- Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
- Ondokuz Mayıs Üniversitesi Adresli: Evet
Özet
Numerous mathematical expressions for growth models have been developed, but each has its own characteristics and limitations. Therefore, this study has investigated whether artificial intelligence (AI) methods can be an alternative to these models. To this aim, four nonlinear (NL) models (logistic, Richards, Gompertz-Laird, and von Bertalanffy) and three AI techniques — artificial neural networks (ANN), integrated adaptive neuro-fuzzy inference systems with grid partitioning and subtractive clustering (ANFIS-GP and ANFIS-SC) — were used to analyze growth. Some statistical methods, including the mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE) were used to evaluate the model performance. As a result of the study, it was determined that the ANFIS-SC model yielded a better fit with the broiler data due to its low MAE, RMSE, and MAPE values (7.68 g, 11.93 g, and 1.06%, respectively). The overall recommendation of this study is that the AI models could be used as an alternative to determine a broiler growth curve.