Search

This paper presents a multi-algorithm fusion model (StackingGroup) based on the Stacking ensemble learning framework to address the variable selection problem in high-dimensional group structure data. The proposed algorithm takes into account the differences in data observation and training principles of different algorithms. It leverages the strengths of each model and incorporates Stacking ensemble learning with multiple group structure regularization methods. The main approach involves dividing the data set into K parts on average, using more than 10 algorithms as basic learning models, and selecting the base learner based on low correlation, strong prediction ability, and small model error. Finally, we selected the grSubset + grLasso, grLasso, and grSCAD algorithms as the base learners for the Stacking algorithm. The Lasso algorithm was used as the meta-learner to create a comprehensive algorithm called StackingGroup. This algorithm is designed to handle high-dimensional group structure data. Simulation experiments showed that the proposed method outperformed other R2, RMSE, and MAE prediction methods. Lastly, we applied the proposed algorithm to investigate the risk factors of low birth weight in infants and young children. The final results demonstrate that the proposed method achieves a mean absolute error (MAE) of 0.508 and a root mean square error (RMSE) of 0.668. The obtained values are smaller compared to those obtained from a single model, indicating that the proposed method surpasses other algorithms in terms of prediction accuracy. [ABSTRACT FROM AUTHOR]

A simple, sensitive, and accurate stability-indicating analytical method has been developed and validated using ultra high performance liquid chromatography. The developed method is used to evaluate the related substances of eplerenone (EP). The degradation behavior of EP under stress conditions was determined, and the major degradants were identified by ultra high performance liquid chromatography with tandem mass spectrometry. The chromatographic conditions were optimized using an impurity-spiked solution, and the samples, generated from forced degradation studies. The resolution of EP, its potential impurities, and its degradation products was performed on a Waters UPLC BEH C18 column (50 × 2.1 mm, 1.7 μm) by linear gradient elution using a mobile phase consisting of 10 mmol/L ammonium acetate adjusted to pH 4.5, methanol and acetonitrile. A photo-diode array detector set at 245 nm was used for detection. The flow rate was set at 0.3 mL/min. The procedure had good specificity, linearity (0.02-3.14 μg/mL), recovery (96.1-103.9%), limit of detection (0.01-0.02 μg/mL), limit of quantitation (0.03-0.05 μg/mL), and robustness. The correction factors of the process-related substances were calculated. [ABSTRACT FROM AUTHOR]

Abstract: Aqueous-phase reforming (APR) of the low-boiling fraction (LBF) of bio-oil from rice husk pyrolysis was investigated over Pt/Al2O3 for hydrogen production. The influence of reaction temperature (503–563K), reaction time (1–4h), different load of organics in water (3–12wt.%) and catalyst amounts (0.25–1g) were studied. The hydrogen content of the gas products reached 65vol.% at 533K for 4h with a feedstock of 9.6g LBF and 30mL water in the presence of 0.5g of 2wt.% Pt/Al2O3. APR is a promising method to make use of the organics in LBF of bio-oil to produce hydrogen. [Copyright &y& Elsevier]

Abstract: In this work, the upgrading reactions of high-boiling fraction (HBF) of bio-oil were carried out over a series of supported mono- and bi-metallic catalysts under the supercritical methanol condition. During these reactions, esterification and cracking (alcoholysis and hydrocracking) were the two dominant processes. PtNi/MgO exhibited good performance, and gave a high yield (72.4wt.%) of refined oil. The acid–base properties of the supports have an important effect on the coke deposition on the catalyst surface. The acidic catalysts gave the somewhat lower product yields, but tended to inhibit coking reaction. This would improve the life of the catalysts in the practical applications. The refined oil is believed to be a potential substitute or partial substitute for the fossil transportation fuel. [Copyright &y& Elsevier]

Abstract: In this work, the low-boiling fraction (LBF) of bio-oil was used as feed stock. LBF is a very complex mixture, and the three groups in LBF: acids, aldehydes and phenols, are primarily responsible for deterioration in the quality. The upgrading reactions were carried out over Pt/Al2(SiO3)3, Pt/C or Pt/MgO in supercritical methanol. It is demonstrated that supercritical condition can greatly facilitate the esterification process, and after 6h reaction, all the acids can be converted into esters even without adding any catalyst. The total amount of the three groups left in products was much less exhibited on Pt supported on active carbon and MgO in the presence of hydrogen. By investigating the model reactions, the relations between the representative compounds and major products were identified, and the conversion scheme of the upgrading reactions is proposed. [Copyright &y& Elsevier]

Abstract: A study of the liquid-phase conversion of biomass-derived sugars directly to methyl lactate catalyzed over solid base catalysts is presented. The yield of methyl lactate reached 29.45% under optimal reaction conditions. The effects of reaction time, reaction temperature, and catalysts calcination temperature were investigated. Fresh and used catalysts were characterized by XRD, CO2-TPD, and N2-adsorption techniques. MgO with the highest concentration of weakly basic sites was found to show the highest catalytic activity for the conversion of sugars to methyl lactate. [Copyright &y& Elsevier]