Prediction of overall yield of Gynura procumbens from ethanol-water + supercritical CO2 extraction using artificial neural network model
Gynura procumbens, also known as Sambung nyawa by Malays, is a traditional herb that can be found in Malaysian forests. Traditional Malays rely on this plant in many ways. The leaves can be eaten raw as a salad, and it is believed to have medicinal effects for humans suffering from cancer, diabetes,...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier Ltd
2022
|
Series: | Case Studies in Chemical and Environmental Engineering
|
Subjects: | |
Online Access: | View Fulltext in Publisher View in Scopus |
Summary: | Gynura procumbens, also known as Sambung nyawa by Malays, is a traditional herb that can be found in Malaysian forests. Traditional Malays rely on this plant in many ways. The leaves can be eaten raw as a salad, and it is believed to have medicinal effects for humans suffering from cancer, diabetes, and other afflictions. The extraction of G. procumbens has been conducted using various techniques, including supercritical fluid extraction (SFE). The study presents the methodology for an artificial neural network (ANN) model with the intention of predicting the overall yield, YT of G. procumbens in supercritical fluid depending on the SFE parameter, namely, water content in ethanol (10–30% v/v), temperature (60–70 °C), and pressure (18–24 MPa). The overall yield from SFE was optimal with a 15.90% wt/wt sample at 30% water content in ethanol, 70 °C and 24 MPa. A three-layer ANN model was generated using the experimental data obtained. The Levenberg–Marquardt back-propagation (LMP) algorithm with a trainlm transfer function was used to predict the YT. A total of 1–8 neurons in the hidden layer were trained, and the most suitable network found was with 6 neurons in the hidden layer. The value of the mean square error (MSE) was 0.3011, and the coefficient of determination (R2) was 0.9877, showing that the (3-6-1) model is a better option for predicting the YT. © 2021 The Authors |
---|---|
ISBN: | 26660164 (ISSN) |
DOI: | 10.1016/j.cscee.2021.100175 |