疫情下电磁场理论课程的网络教学

本文主要总结了新冠疫情期间作者的电磁场理论课程在线教学经验。对比分析了录播和直播的优缺点后,选择录播教学方式。基于超星网络教学平台,展示了录播网络教学的具体措施,包括网上答疑和学习效果检查以及在线批改作业等。给出了网络教学可以为线下教学继续使用的方法和手段,为疫情结束后的正常教学提供了新的网络教学补充措施。     

Hyperparameter on-line learning of stochastic resonance based threshold networks

Aiming at training the feed-forward threshold neural network consisting of nondifferentiable activation functions, the approach of noise injection forms a stochastic resonance based threshold network that can be optimized by various gradient-based optimizers. The introduction of injected noise extends the noise level into the parameter space of the designed threshold network, but leads to a highly non-convex optimization landscape of the loss function. Thus, the hyperparameter on-line learning procedure with respective to network weights and noise levels becomes of challenge. It is shown that the Adam optimizer, as an adaptive variant of stochastic gradient descent, manifests its superior learning ability in training the stochastic resonance based threshold network effectively. Experimental results demonstrate the significant improvement of performance of the designed threshold network trained by the Adam optimizer for function approximation and image classification.     

Learn decision trees with deep visual primitives

In this paper, we strive to propose a self-interpretable framework, termed PrimitiveTree, that incorporates deep visual primitives condensed from deep features with a conventional decision tree, bridging the gap between deep features extracted from deep neural networks (DNNs) and trees’ transparent decision-making processes. Specifically, we utilize a codebook, which embeds the continuous deep features into a finite discrete space (deep visual primitives) to distill the most common semantic information. The decision tree adopts the spatial location information and the mapped primitives to present the decision-making process of the deep features in a tree hierarchy. Moreover, the trained interpretable PrimitiveTree can inversely explain the constituents of the deep features, highlighting the most critical and semantic-rich image patches attributing to the final predictions of the given DNN. Extensive experiments and visualization results validate the effectiveness and interpretability of our method.     

Cyberbullying: Effect of emergency perception on the helping tendencies of bystanders

The effect of the emergency perception of bystanders of cyberbullying victims on helping behaviors is often neglected in research on cyberbullying. In this study, we explored the influence of this cognitive factor on cyber-bystanders’ helping tendencies as well as elucidated possible underlying processes. The results of two studies were reported. In Study 1, 150 undergraduates read a true case of a girl experiencing cyberbullying. The results indicated that when the participants perceived the victim’s situation to be more critical (i.e., higher emergency perception), their helping tendencies were stronger, partly through increased state empathy followed by feelings of responsibility to help. In Study 2, we randomly assigned 300 undergraduates to two groups. The low emergency group read the same cyberbullying case as Study 1, whereas the cyberbullying case read by the high emergency group contained additional emergency information of the victim. The results indicated that the high emergency group expressed stronger helping tendencies than did the low emergency group. This effect was caused by a stronger perception that the victim was in an emergency situation, which not only strengthened the participants’ helping tendencies directly but also indirectly through increasing their state empathy and feelings of responsibility to help.     

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.     

Prediction of the drag reduction effect of pulsating pipe flow based on machine learning

Prediction of drag reduction effect caused by pulsating pipe flows is examined using machine learning. First, a large set of flow field data is obtained experimentally by measuring turbulent pipe flows with various pulsation patterns. Consequently, more than 7000 waveforms are applied, obtaining a maximum drag reduction rate and maximum energy saving rate of 38.6% and 31.4%, respectively. The results indicate that the pulsating flow effect can be characterized by the pulsation period and pressure gradient during acceleration and deceleration. Subsequently, two machine learning models are tested to predict the drag reduction rate. The results confirm that the machine learning model developed for predicting the time variation of the flow velocity and differential pressure with respect to the pump voltage can accurately predict the nonlinearity of pressure gradients. Therefore, using this model, the drag reduction effect can be estimated with high accuracy.     

Metabolomics techniques applied in the investigation of phenolic acids from the agro-industrial by-product of Carapa guianensis Aubl

Processing of Carapa guianensis seeds to obtain oil on an industrial scale generates a significant amount of by-product, approximately 66% w/w, which is called cake and is a potential source of biomolecules, including simple phenolic structures. For this reason, studies were carried out on the chemical profiles of hydrolyzed extract from this agro-industrial by-product through High Performance Thin-Layer Chromatography (HPTLC) and Gas Chromatography coupled to Mass Spectrometry (GC–MS). These techniques were used to detect metabolic classes and/or groups, and to identify, for the first time, thirteen simple phenolic acids in this by-product. The sample antioxidant capacity was determined by methods of 2,2-diphenyl-1-picrylhydrazyl (DPPH)and 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS⁺) radicals direct sequestration. The hydrolyzed fraction showed a total of 63.47% in the relative abundance of the total of compounds, standing out: p-hydroxybenzoic acid (39.19%) and protocatechuic acid (3,4-dihydroxybenzoic acid) (5.62%), both from hydroxybenzoic acids and 3-(3,4-dihydroxyphenyl)lactic acid, (7.76%) hydroxycinnamic acids derivatives. In these results, the fraction rich in simple phenolic acids was obtained, attributing the prominent behavior of this matrix antioxidant activity, expressed by (IC₅₀: of 16.42 µg/mL and 6.52 µg/mL for DPPH and ABTS⁺ radicals, respectively). The research demonstrated an alternative to applicability that involves sustainability from agro-industrial. These techniques were used to detect metabolic classes and/or groups, and to identify, for the first time, thirteen simple phenolic acids in this by-product, generating a process capable of converting biomass into a bioproduct, consisting of bioactive compounds, in addition to adding value to the industrial chain.     

紫色球杆菌视紫红质光谱特性的机器学习研究

近年来，机器学习等人工智能技术被应用于蛋白质工程，其在蛋白质结构、功能预测、催化活性等研究中具有独特优势。在未知蛋白质结构的情况下，将蛋白质序列和功能特性与机器学习相结合，基于序列-活性关系（innovative sequence-activity relationship，ISAR）算法，将蛋白质氨基酸序列数字化，用快速傅里叶变换（fast four transform，FFT）进行预处理，再进行偏最小二乘回归建模，可在数据集较少情况下拟合得到最佳模型。通过机器学习对紫色球杆菌视紫红质（gloeobacter violaceus rhodopsin，GR）的突变体蛋白质氨基酸序列与光谱最大吸收波长进行建模，获得了最佳模型。用最佳索引LEVM760106建模得到的确定系数R² 为0.944，均方误差E为11.64。用小波变换进行的预处理，其R² 虽也约为0.944，但E大于11.64，不及FFT进行的预处理。方法较好地解决了蛋白质序列与功能特性之间的数学建模问题，在蛋白质工程中可为预测更优的突变体提供支持。     

基于串联两阶段网络DEA模型的供应链规模收益分析

数据包络分析(DEA)是评价供应链系统(Supply chain system)间相对有效性的一种重要的工具,但是传统的DEA不考虑供应链的内部结构,对系统效率评价偏高;而本文所研究两阶段串联供应链系统,考虑把部分中间产品作为最终产品输出,增加额外中间投入的情形.基于所提出的供应链系统结构,本文建立相应的串联结构下的网络DEA模型,并针对所建立模型进行相关理论的研究,给出了串联结构下的生产可能集和规模收益情况判定方法.最后,进行数值实验,以验证我们提出的结论.     

Sobolev Stability of Plane Wave Solutions to the Nonlinear Schrödinger Equation

This article explores the questions of long time orbital stability in high order Sobolev norms of plane wave solutions to the NLSE in the defocusing case.