基于GA_XGBoost模型的大学生科研能力预测问题研究

科学评价大学生科研创新能力对我国科研水平的提高具有重要意义.采用机器学习模型来预测大学生科研能力可以起到良好的效果,提出一种GAXGBoost模型来实现对大学生的科研能力预测.此模型是以Xgboost算法为基础,然后充分利用遗传算法的全局搜索能力自动搜索Xgboost最优超参数,避免了人为经验调参不准确的缺陷,最后采用精英选择策略以此确保每一轮都是最佳的进化结果.通过分析表明,所采用的GAXGBoost模型在大学生科研能力预测的结果中具有很高的精度,将此模型与Logistic Regression、Random Forest、SVM等模型进行对比,GAXGBoost模型的预测精度最高.     

Optimal control and numerical methods for hybrid stochastic SIS models

This work focuses on optimal controls of a class of stochastic SIS epidemic models under regime switching. By assuming that a decision maker can influence the infectivity period, our aim is to minimize the expected discounted cost due to illness, medical treatment, and the adverse effect on the society. In addition, a model with the incorporation of vaccination is proposed. Numerical schemes are developed by approximating the continuous-time dynamics using Markov chain approximation methods. It is demonstrated that the approximation schemes converge to the optimal strategy as the mesh size goes to zero. Numerical examples are provided to illustrate our results.     

Negative thermal expansion in NbF_3 and NbOF_2:A comparative theoretical study

Thermal expansion control is always an obstructive factor and challenging in high precision engineering field. Here,the negative thermal expansion of Nb F_3 and Nb OF_2 was predicted by first-principles calculation with density functional theory and the quasi-harmonic approximation(QHA). We studied the total charge density, thermal vibration, and lattice dynamic to investigate the thermal expansion mechanism. We found that the presence of O induced the relatively strong covalent bond in Nb OF_2, thus weakening the transverse vibration of F and O in Nb OF_2, compared with the case of Nb F_3.In this study, we proposed a way to tailor negative thermal expansion of metal fluorides by introducing the oxygen atoms.The present work not only predicts two NTE compounds, but also provides an insight on thermal expansion control by designing chemical bond type.     

基于模糊神经元PID算法的注塑机注射速度控制算法研究

基于一款市场较为畅销的注塑机, 设计出一种能精确控制注射速度的模糊神经元PID控制器. 首先, 设计出具有自学能力的神经元PID控制器, 利用模糊算法对其进行优化; 其次, 在原有注射速度线性数学模型的基础上, 构建注塑机注射速度的非线性模型; 最后, 利用MATLAB在所建数学模型的基础上对模糊神经元PID控制器进行仿真实验. 实验结果表明, 所设计控制器具有响应迅速、无超调量、控制精度高、控制稳定等优点.     

富晶格位错的多孔铋纳米花高效电还原二氧化碳制甲酸盐※

二氧化碳转化已成为现今世界研究的热点. 本工作采用原位电化学转化的策略, 将简单溶剂热法合成的层状甲酸氧铋纳米花(BiOCOOH NFs)还原为带有大量晶格位错的多孔铋纳米花(p-Bi NFs). 研究结果表明, p-Bi NFs电催化二氧化碳转化为甲酸盐具有较小的过电位(436 mV). 在–1.8 V(相对饱和甘汞电极, vs. SCE)时, 甲酸盐的分电流密度(j_formate)高达24.4 mA•cm^-2, 法拉第效率(FE_formate)为96.7%, 且在超过500 mV的宽电位窗口内FE_formate超过90%, 并具有很好的稳定性. 该催化剂的高催化性能可归因于前驱体晶格坍塌和重构而形成特殊的多孔粗糙的微纳多级结构, 其表面富含晶格位错和缺陷等高本征活性位, 且具有较强的电子传递能力. 本研究为设计合成高性能的电催化二氧化碳还原产甲酸催化剂提供了新的思路.     

Pure annihilation decay of strange beauty meson into two charm heavy mesons

In this study, the heavy to heavy decay of \begin{document}$ B^0_s\rightarrow D^{*+}D^- $\end{document} is evaluated through the factorization approach by using the final state interaction as an effective correction. Under the factorization approach, this decay mode occurs only through the annihilation process, so a small amount is produced. Feynman's rules state that six meson pairs can be assumed for the intermediate states before the final meson pairs are produced. By taking into account the effects of twelve final state interaction diagrams in the calculations, a significant correction is obtained. These effects correct the value of the branching ratio obtained by the pure factorization approach from \begin{document}$ (2.41\pm1.37)\times10^{-5} $\end{document} to \begin{document}$ (8.27\pm2.23)\times10^{-5} $\end{document}. The value obtained for the branching ratio of the \begin{document}$ B^0_s\rightarrow D^{*+}D^- $\end{document} decay is consistent with the experimental results.     

A novel UPLC–MS/MS method for simultaneous quantification of trigonelline, 4-hydroxyisoleucine,and diosgenin from Trigonella foenum-graecum extract: Application to pharmacokinetic study in healthy and type 2 diabetic rats

Trigonelline (TR), 4-hydroxyisoleucine (4-HI), and diosgenin (DG) are the main bioactives of the purified standardized extract of the popular plant Trigonella foenum-graecum L. (TFG), and it has been proven effective for the treatment of various diseases. However, to the best of our knowledge, no study has investigated the pharmacokinetic parameters of purified standardized T. foenum-graecum extract in normal and diabetic Wistar rats. The present study has developed and validated a rapid, reliable, and sensitive simultaneous ultra-performance liquid chromatography MS method to estimate these bioactives. The chromatographic separation was achieved using methanol, acetonitrile, and 0.1% formic acid with the ideal gradient flow system on a BEH Shield RP 18 column. A positive electrospray ionization mode was selected to estimate m/z values of TR (138.14 > 94.63), 4-HI (148.19 > 74.08), and DG (415.54 > 271.33). The method was robust and reproducible over the linearity range of 60–5000, 6–5000, and 15–5000 ng/mL for TR, 4-HI, and DG, respectively. Using this novel validated method, we investigated the pharmacokinetic parameters of bioactives using Phoenix WinNonlin version 8.0 (Certera) in normal and diabetic rats. The assay was successfully applied for the estimation of pharmacokinetic parameters using noncompartmental analysis. This investigation shows that the absorption rate increased, whereas distribution and elimination processes slowed down in diabetic rats compared with normal rats.