Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-New research on Technology is the subj ect of a report. According to news reporting from Beijing, People's Republic of China, by NewsRx journalists, research stated, "In the quest for enhanced precis ion in near-infrared spectroscopy (NIRS), in this study, the application of a no vel BEST-1DConvNet model for quantitative analysis is investigated against conve ntional support vector machine (SVM) approaches with preprocessing such as multi plicative scatter correction (MSC) and standard normal variate (SNV). We assesse d the performance of these methods on NIRS datasets of diesel, gasoline, and mil k using a Fourier Transform Near-Infrared (FT-NIR) spectrometer having a wavelen gth range of 900-1700 nm for diesel and gasoline and 4000-10,000 nm for milk, en suring comprehensive spectral capture." The news correspondents obtained a quote from the research from the Beijing Inst itute of Petrochemical Technology, "The BEST-1DConvNet's effectiveness in chemom etric predictions was quantitatively gauged by improvements in the coefficient o f determination (R2) and reductions in the root mean square error (RMSE). The BE ST-1DConvNet model achieved significant performance enhancements compared to the MSC + SNV + 1D + SVM model. Notably, the R2 value for diesel increased by appro ximately 48.85% despite a marginal RMSE decrease of 0.92% . R2 increased by 11.30% with a 3.32% RMSE reduction for gasoline, and it increased by 8.71%, accompanied by a 3.51% RMSE decrease for milk."
NETL
NSTL