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.     

Comparison study of exchange-correlation functionals on prediction of ground states and structural properties

Despite significant advances in first-principles calculation methods, there is no single exchange-correlation functional which predicts the ground state of materials without an error yet. We investigated how accurately ground states of binary semiconductors are described using 16 exchange-correlation functionals (with or without van der Waals corrections). LDA, PBEsol, SCAN (with or without rVV10 correction), and PBE with D3 van der Waals correction (zero or Becke-Johnson damping) show good predicting power. The lattice constants of stable phases were slightly better described by SCAN, PBEsol, PBE+D3 (Becke-Johnson damping), and MS2. We also propose a set of functionals to double-check the stability of new materials based on the majority vote.     

基于不确定理论的铁路货运需求预测

合理预测铁路货运需求是铁路管理部门建设、运营等决策基础。为应对铁路货运需求的复杂变化,基于Pearson相关性分析方法筛选出铁路货运需求的七个具有关键影响的因素,并结合不确定理论建立不确定多元线性回归模型,相应的铁路货运预测结果由传统单一值变成可能的需求区间范围,更加符合处于不确定环境下的铁路货运需求实际情况。选取国家统计局2004~2016年相关数据进实证研究,并与回归模型以及BP模型的预测结果对比分析,实验表明不确定多元线性回归的预测结果更加精确。     

Identifying aspirin polymorphs from combined DFT-based crystal structure prediction and solid-state NMR

A combined experimental and computational approach was used to distinguish between different polymorphs of the pharmaceutical drug aspirin. This method involves the use of ab initio random structure searching (AIRSS), a density functional theory (DFT)-based crystal structure prediction method for the high-accuracy prediction of polymorphic structures, with DFT calculations of nuclear magnetic resonance (NMR) parameters and solid-state NMR experiments at natural abundance. AIRSS was used to predict the crystal structures of form-I and form-II of aspirin. The root-mean-square deviation between experimental and calculated ¹H chemical shifts was used to identify form-I as the polymorph present in the experimental sample, the selection being successful despite the large similarities between the molecular environments in the crystals of the two polymorphs.     

A Dynamic Anomaly Detection Approach Based on Permutation Entropy for Predicting Aging-Related Failures

Software aging is a phenomenon referring to the performance degradation of a long-running software system. This phenomenon is an accumulative process during execution, which will gradually lead the system from a normal state to a failure-prone state. It is a crucial challenge for system reliability to predict the Aging-Related Failures (ARFs) accurately. In this paper, permutation entropy (PE) is modified to Multidimensional Multi-scale Permutation Entropy (MMPE) as a novel aging indicator to detect performance anomalies, since MMPE is sensitive to dynamic state changes. An experiment is set on the distributed database system Voldemort, and MMPE is calculated based on the collected performance metrics during execution. Finally, based on MMPE, a failure prediction model using the machine learning method to reveal the anomalies is presented, which can predict failures with high accuracy.     

Fatigue Behavior of Composite Sandwich Panels Under Three Point Bending Load

In this paper, the bending fatigue tests of honeycomb sandwich panels are carried out by using an improved three-point bending test fixture, and the S-N curves at different stress ratios are obtained. Through the records of fatigue damage in the experiment, the failure mode of the honeycomb sandwich panels and the source of fatigue damage are determined. At the same time, through the calculation of the shear stress distribution on the honeycomb wall, the reasons for the difference in the failure morphology of the L-direction and W-direction sandwich panels are clarified. Besides, a life prediction method is proposed and its effectiveness in predicting the fatigue life of sandwich panels has been verified.     

基于SVM的固化土无侧限抗压强度模型

为实现较少试验次数下固化土无侧限抗压强度(qu)的准确预测, 提出了基于支持向量机(SVM)的固化土qu的预测模型. 以固化剂各组分掺入比、龄期、初始含水量、固化剂掺量等因素为输入量, 固化土的qu作为输出量, 以径向基为核函数, 采用网格搜索法和交叉验证法进行参数优化, 建立了基于SVM的固化土qu的预测模型. 算例分析表明: 该模型适用于任意条件下固化土qu的精确预测, 且在较小试验成本下实现与响应面法相当的预测精度.     

一种经验型疲劳小裂纹扩展模型

金属材料疲劳寿命由裂纹萌生和裂纹扩展寿命两部分组成，其中对于萌生寿命中的小裂纹分析是精确描述裂纹萌生寿命的关键．而小裂纹在扩展过程中由于尺寸相对较小，导致传统线弹性断裂力学预测方法失效，需要对其进行改进，考虑裂纹尖端塑性区引起的残余压应力对小裂纹扩展速度的影响．本文针对此问题进行了初步分析，通过对塑性区引起的残余应力的量化，结合小裂纹门槛值特性，提出了一种经验型修正的小裂纹扩展模型，用于定量预测裂纹的萌生寿命．使用铝合金6082-T6缺口试样进行了疲劳实验，并与理论结果进行了对比，验证了所提模型的有效性．     

Reducing the computational cost of NMR crystallography of organic powders at natural isotopic abundance with the help of 13C-13C dipolar couplings

Structure determination of functional organic compounds remains a formidable challenge when the sample exists as a powder. Nuclear magnetic resonance crystallography approaches based on the comparison of experimental and Density Functional Theory (DFT)-computed ¹H chemical shifts have already demonstrated great potential for structure determination of organic powders, but limitations still persist. In this study, we discuss the possibility of using ¹³C-¹³C dipolar couplings quantified on powdered theophylline at natural isotopic abundance with the help of dynamic nuclear polarization, to realize a DFT-free, rapid screening of a pool of structures predicted by ab initio random structure search. We show that although ¹³C-¹³C dipolar couplings can identify structures possessing long range structural motifs and unit cell parameters close to those of the true structure, it must be complemented with other data to recover information about the presence and the chemical nature of the supramolecular interactions.     

融合图卷积神经网络和注意力机制的PM2.5小时浓度多步预测

PM_2.5小时浓度多为单步预测。为实现PM_2.5小时浓度的多步预测，基于“编码器-解码器”的序列-序列预测（Seq2Seq）模型，集合图卷积神经网络提取非欧式空间数据特征的能力以及注意力机制自适应关注特征的能力，提出了融合图卷积神经网络和注意力机制的PM_2.5小时浓度多步预测（GCN_Attention_Seq2Seq）模型。并与Seq2Seq模型和使用了图卷积神经网络、未使用注意力机制的GCN_Seq2Seq模型进行了对照，以2015—2016年北京市22个空气质量监测站点的空气质量数据为样本进行实例验证，结果表明，Seq2Seq模型和图卷积神经网络（GCN）可对PM_2.5小时浓度数据的时空依赖进行有效建模，注意力机制有助于减缓多步预测中的预测精度衰减，提升PM_2.5小时浓度多步预测的精度。GCN_Attention_Seq2Seq模型可有效应用于多种长度的PM_2.5浓度预测窗口。