Abstract
2024 OCT 09 (NewsRx)-By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News-New research on Machine Learning is the subject o f a report. According to news reporting out of Anhui, People's Republic of China , by NewsRx editors, research stated, "Cellulose and hemicellulose are key cross -linked carbohydrates affecting bioethanol production in maize stalks. Tradition al wet chemical methods for their detection are labor-intensive, highlighting th e need for high-throughput techniques." Our news journalists obtained a quote from the research from Anhui Agricultural University, "This study used Fourier transform infrared (FTIR) spectroscopy comb ined with machine learning (ML) algorithms on 200 large-scale maize germplasms t o develop robust predictive models for stalk cellulose, hemicellulose and holoce llulose content. We identified several peak height features correlated with thre e contents, used them as input data for model building. Four ML algorithms demon strated higher predictive accuracy, achieving coefficient of determination ® ran ging from 0.83 to 0.97. Notably, the Categorical Boosting algorithm yielded opti mal models with coefficient of determination ® exceeding 0.91 for the training s et and over 0.81 for the test set."
NETL
NSTL