Feature validity during machine learning paradigms for predicting biodiesel purity

The main effort of this study is to examine the feasibility of four novel machine learning models namely Alternating Model Tree, Random Tree, Least Median Square, and Multi-Layer Perceptron Regressor to estimate the biodiesel purity. Then, the mentioned methods are utilized to identify a relationship between the input and output parameters of the biodiesel system. The parameter response was taken as the essential output of fatty acid methyl ester, while the input parameters opted the oil type, catalyst type, catalyst concentration, reaction temperature, methanol-to-oil ratio, reaction time, frequency as well as amplitude. The predicted results obtained by the tools mentioned supra were evaluated according to several known statistical indices. The obtained results proved that the AMT is the best predictive network.