食品中兴奋剂污染研究现状

随着食品和膳食补充剂的市场变得越来越全球化,食品和膳食补充剂的安全性、质量和功效引起人们的高度关注。近年来,食品和膳食补充剂中被检测出兴奋剂阳性的事件屡见不鲜。运动员在误服误用被兴奋剂污染的食品和膳食补充剂后,会导致兴奋剂检测呈阳性,这对运动员和国家都造成了重大损失。由于摄入受污染的食品或膳食补充剂会导致严重的健康损害或意外违反反兴奋剂规定,因此准确了解食品和膳食补充剂中兴奋剂污染种类是十分有必要的。本文主要从食品和膳食补充剂中兴奋剂污染的来源和种类以及常用的检测方法等方面进行简要概述,以提高运动员对高风险食品的警惕和防范,避免因误服被兴奋剂污染的食品、膳食补充剂而导致的不良分析结果。     

成分配比对稀土钨电极材料组织及力学性能影响

通过固液掺杂、等静压压制、中频烧结的方法，制备了不同的氧化镧、氧化钇、氧化锆三元掺杂成分比例的钨电极材料烧结棒材，探究了不同成分配比对样品显微组织、第二相粒子分布以及宏观力学性能的影响。结果表明，氧化镧、氧化钇、氧化锆三元复合添加能够有效改善第二相粒子在钨基体中的分布形态，降低第二相在晶界的过度富集，提高钨电极材料的综合力学性能。并且当添加成分镧、钇、锆质量比为3:1:1时，材料具有最好的综合力学性能，致密度可达96.04%，显微硬度可达549.37HV0.3，抗压强度可达3785MPa，原因是此配比下第二相粒子最为细小均匀，弥散程度最高，对基体晶粒的细化作用最好，该配比下钨基体平均晶粒尺寸达到10.3μm。     

液相色谱-串联质谱技术在临床检验中的应用研究进展

液相色谱-串联质谱（LC-MS/MS）技术具有高灵敏度、高特异性、高分辨率和高效率的优点。近年来随着仪器灵敏度的提高，LC-MS/MS在常规临床检验中显示出极大的潜力，并在疾病早期预防和诊断中发挥着不可替代的作用。本文对LC-MS/MS在新生儿疾病筛查、维生素D检测、内分泌激素检测、肽类和蛋白质定量分析等临床检验方面的研究进展进行综述，并讨论了未来面临的挑战。     

Facial-sketch Synthesis: A New Challenge

Machine Intelligence Research - This paper aims to conduct a comprehensive study on facial-sketch synthesis (FSS). However, due to the high cost of obtaining hand-drawn sketch datasets, there is a...     

基于VMD和BSA-KELM的高陡边坡位移预测模型研究

边坡位移的时间序列曲线存在复杂的非线性特性,传统的预测模型精度不足以满足预测要求。为此提出了基于变分模态分解的鸟群优化-核极限学习机的预测模型,并用于河北省某水泥厂的边坡位移预测。该方法首先采用VMD把边坡位移序列分解为一系列的有限带宽的子序列,再对各子序列分别采用相空间重构并用核极限学习机预测,采用鸟群算法优化相空间重构的嵌入维度和KELM中惩罚系数和核参数三个数值,以取得最优预测模型。最后将各个子序列预测值叠加,得到边坡位移的最终预测值。结果表明：和KELM、BSA-KELM、EEMD-BSA-KELM模型相比,基于VMD的BSA-KELM预测精度更高,为边坡位移的预测提供一种有效的方法。     

Comparison of yield and metabolites according to the types of hilling materials utilized during Glehnia littoralis sprout vegetable cultivation

Food Science and Biotechnology - Various hilling materials (rice hulls, pine sawdust, and perlite) were compared to produce sprout vegetables using beach silvertop (Glehnia littoralis Fr. Schm. ex...     

Evaluating differentially private decision tree model over model inversion attack

International Journal of Information Security - Machine learning techniques have been widely used and shown remarkable performance in various fields. Along with the widespread utilization of...     

基于1D CNN与XGBoost的恶意代码纹理检测

恶意代码数量已经呈现爆炸式增长,对于恶意代码的检测防护显得尤为重要.近几年,基于深度学习的恶意代码检测方法开始出现,基于此,提出一种新的检测方法,将恶意代码二进制文件转化为十进制数组,并利用一维卷积神经网络(1 Dimention Convolutional Neural Networks,1D CNN)对数组进行分类和识别.针对代码家族之间数量不平衡的现象,该算法选择在分类预测上表现良好的XGBoost,并对Vision Research Lab中的25个不同恶意软件家族的9458个恶意软件样本进行了实验.实验结果表明,所提的方法分类预测精度达到了97％.     

Facile preparation and strong adhesive strength of honeycomb polyurethane films with small pore diameter

With the continuous development of bionics, such as, geckos and virginia creeper with both superhydrophobic and super-adhesive, the surface wetting and super-adhesive properties of various porous materials have attracted extensive attention of the scientific and medical communities. Here, the honeycomb polyurethane (PU) porous films with strong adhesion were successfully prepared by microphase separation method and the effects of growth parameters on their microstructure and adhesive strength to ice were investigated. It was found that a high relative humidity (e.g., 100%) and a low solution concentration (e.g., 2%) facilitated the formation of ordered honeycomb PU porous films, and as-prepared PU pores with average pore diameter as small as 5 μm are better ordered and more uniform than these in related documents. Although the contact angle of water droplets on the surface of PU porous films increased from the premodification value of 85–130° to more than 160° after surface modification with polydopamine (PDA), the corresponding rolling angle remained approximately constant (180°), indicating that the surface of PU porous films has strong adhesion similar to geckos and virginia creeper. Furthermore, at lower temperature, the PU porous films exhibited the high adhesive strength of 142.13 kPa on ice, which was strongly dependent on the porous microstructures and surface compositions. The improved adhesive behavior to ice of honeycomb PU porous films modified with PDA provides new strategies for surface modification of materials and potential applications in medical domain.