Artificial Neural Network Predictive Modelling of luffa cylindrica Seed Oil Antioxidant Yield

Abstract

This study applied artificial neural network (ANN) in evaluating the models for terpineol and polyphenol yield from luffa cylindrica seed oil. The experiment was carried out at a temperature (60-80 o C), time (4-6 hours), and solvent/seed ratio (8-12 ml/g) with response as antioxidant yield. FTIR (Fourier Transform Infra-red Spectroscopy) revealed the presence of terpineol and polyphenol at peaks of 1461.1cm -1 and 3008.0cm -1 respectively. The ANN prediction indices are thus; terpineol (R 2 = 9.9999E-1, MSE=2.25766E-9) and polyphenol (R 2 =9.9999E-1, MSE=4.42588E-10). This study reveals that the ANN technique can successfully predict antioxidants from luffa cylindrica seed oil.ANN has shown its capability of controlling imprecise relationships among variables (Maosudi et al., 2018); ANN comprises mathematical models that apply biological neurons in solving intricate and uncertain processes, hence can be relied upon than linear and multivariate statistical procedures that have shown their Cite Nwosu-Obieogu, K. ( 2021). Artificial neural network predictive modelling of luffa cylindrica seed oil antioxidant yield. GU J Sci, 8(4), 494-504.