Estimasi Emisi Metana pada Lingkungan Rumen in Vitro menggunakan Artificial Neural Network

Abstract

The production of methane from ruminants contributes significantly for the total methane emissions in the atmosphere. Methane is one of the greenhouse gases that causes global warming. Its ability to retain heat is of 20 times bigger than that of carbon dioxide. In this research, Artificial Neural Networks (ANNs) are used to estimate the methane emissions from volatile fatty acid (vfa) composition in the rumen in vitro; acetate (C2), propionate (C3), butyrate (C4), iso-C4, (valerate) C5, and iso-C5. In this research, after careful tuning, we obtained the best ANNs with the NMSE of 0.0182. This suggests that ANN outperform the stoichiometric equation in ability to estimate methane emissions in the rumen