Application of machine learning in optimization of bioethanol production from lignocellulosic biomass

Abstract

Regardless of the areas, artificial intelligence and machine learning techniques were utilized to analyse a significant amount of data. The two main methods utilized in machine learning for processing a huge amount of data were classification and regression. To estimate the relationship between the data and aid in forecasting the optimum yield, many machine learning techniques will be used to the biomass dataset that is currently provided. Several research works using machine learning algorithm has been carried out in biodiesel production process parameters for estimating biodiesel yield. However, optimization of bioethanol process parameters using machine learning algorithms for estimating and predicting glucose and ethanol yield is a new concept. Comparing the enzymatic hydrolysis process with the dilute acid hydrolysis process, an additional Euclidian distance algorithm for analyzing the importance of process parameters was implemented in optimization to improve the accuracy of the results with a minimum process parameter dataset. For improving the quality of the dataset, an oversampling method is used in this algorithm. newline