Toward Modeling the In Vitro Gas Production Process by Using Propolis Extract Oil Treatment: Machine Learning and Kinetic Models

Abstract

To overcome challenges with in vivo digestibility assessment, in vitro digestibility techniques have been created. The biobased additive in concentrate can effectively promote the in vitro digestibility performance. The primary goal of this study was to evaluate the propolis effect on produced gas by an in vitro procedure in 25−75% proportion of concentrate. Then, machine learning (ML) models such as nonlinear regression techniques and multilayer perceptron neural networks (MLP-NNs) were applied to assess the prediction performance of gas generation during in vitro digestion using propolis treatments with diverse diet components. The MLP-NN was created using 11 nonlinear regression (NLR) and 12 training algorithms. The findings revealed that the logistic−exponential without lag time (LE 0 ) model was chosen as the best nonlinear model for eight diet interventions. Also, the achievements of the MLP-NN model evaluation revealed that the trainbr training procedure with six neurons in the hidden layer can accurately predict the gas output. The prediction errors of the MLP-NN and NLR approaches were not significantly different (R 2 ∼ 0.99). Three-dimensional response surface graphs were drawn with the help of a neural network, and the optimal value was calculated based on it in a simple and intuitive way. This route displayed the optimum propolis treatment, in which the application of 75% propolis ethanol extract in concentrate could significantly increase gas production between 1 and 4 mL/h. The MLP-NN model has more capabilities than NLR in such studies.

Hydrogen Storage on Porous Carbon Adsorbents: Rediscovery by Nature-Derived Algorithms in Random Forest Machine Learning Model

“…A crucial challenge lies in determining whether the knowledge gained from conceptual or experimental research could be utilized to later forecast the characteristics of those other materials having comparable structural and chemical features in a consistent, effective, and informative way. This is the target of the ML technique, which uses patterns in data to be automatically identified [28][29][30]. ML frequently employs methods from data analysis.…”