Experimental study and exergy efficiency prediction of three-leaved yam (<i>Dioscorea dumetorum</i>) starch drying

Abstract

This study predicted three-leaved yam starch drying using adaptive neuro-fuzzy inference system (ANFIS), artificial neural network (ANN) and support vector machine (SVM). The input and output variables were drying temperature, time, air velocity and exergy efficiency (EE) as well as sustainability index (SI), respectively. The EE performance indices are: ANN (R2 = 0.999, MSE = 7.40038E-13), ANFIS (R2 = 0.251, MSE = 0.001) and SVM (R2 = 0.999, MSE = 3.25E-06). The SI prediction indices are: ANN (R2 = 0.99, MSE = 1.0E-10), ANFIS (R2 = 0.41, MSE = 0.010) and SVM (R2 = 0.61, MSE = 0.051). The results indicated that ANN and SVM gave the best prediction for exergy efficiency (EE); while only ANN model is capable of predicting TLYS sustainability index (SI).