基于改进CGANs的入侵检测方法研究

近年来,机器学习算法在入侵检测系统(IDS)中的应用获得越来越多的关注。然而,传统的机器学习算法更多的依赖于已知样本,因此需要尽可能多的数据样本来对模型进行训练。遗憾地是,随着越来越多未知攻击的出现,且用于训练的攻击样本具有不平衡性,传统的机器学习模型会遇到瓶颈。文章提出一种将改进后的条件生成对抗网络(CGANs)与深度神经网络(DNN)相结合的入侵检测模型(CGANs-DNN),通过解决样本不平衡性问题来提高检测模型对未知攻击类型或只有少数攻击样本类型的检测率。深度神经网络(DNN)具有表征数据潜在特征的能力,而经过改进后的条件CGANs,能够通过学习已知攻击样本潜在数据特征分布,来根据指定类型生成新的攻击样本。此外,与生成对抗网络(GANs)和变分自编码器(VAE)等无监督生成模型相比,CGANsDNN经过改进后加入梯度惩罚项,在训练的稳定性上有了很大地提升。通过NSL-KDD数据集对模型进行评估,与传统算法相比CGANs-DNN不仅在整体准确率、召回率和误报率等方面有更好的性能,而且对未知攻击和只有少数样本的攻击类型具有较高的检测率。     

视频侦查中人像智能分析应用及算法优化

人像智能分析指的是对视频或录像中的人像进行结构化和可视化分析,对目标人物进行性别、年龄、发型等特征的智能识别,这项技术在视频侦查中有极高的应用价值。人像识别早期的算法是通过人工提取特征,通过学习低级视觉特征来针对不同属性进行分类学习,这种基于传统方法的模型表现常常不尽如人意。在计算机视觉领域,通过海量图像数据学习的神经网络比传统方法有更丰富的信息量和特征可以被提取。文章尝试通过深度学习技术训练神经网络模型对行人进行检测和识别,对于衣着不同的行人进行智能识别,具有更好的鲁棒性,提升了视频人像识别的准确率,拓展了人工智能技术在身份识别领域的应用。     

Peptide–Peptide Co-Assembly: A Design Strategy for Functional Detection of C-peptide,A Biomarker of Diabetic Neuropathy

Diabetes-related neuropathy is a debilitating condition that may be averted if it can be detected early. One possible way this can be achieved at low cost is to utilise peptides to detect C-peptide, a biomarker of diabetic neuropathy. This depends on peptide-peptide co-assembly, which is currently in a nascent stage of intense study. Instead, we propose a bead-based triple-overlay combinatorial strategy that can preserve inter-residue information during the screening process for a suitable complementary peptide to co-assemble with C-peptide. The screening process commenced with a pentapeptide general library, which revealed histidine to be an essential residue. Further screening with seven tetrapeptide focused libraries led to a table of self-consistent peptide sequences that included tryptophan and lysine at high frequencies. Three complementary nonapeptides (9mer com-peptides), wpkkhfwgq (Trp-D), kwkkhfwgq (Lys-D), and KWKKHFWGQ (Lys-L) (as a negative control) were picked from this table for co-assembly studies with C-peptide. Attenuated total reflectance Fourier transform infrared (ATR-FTIR) and circular dichroism (CD) spectroscopies were utilized to study inter-peptide interactions and changes in secondary structures respectively. ATR-FTIR studies showed that there is indeed inter-peptide interaction between C-peptide and the tryptophan residues of the 9mer com-peptides. CD studies of unaggregated and colloidal C-peptide with the 9mer com-peptides suggest that the extent of co-assembly of C-peptide with Trp-D is greatest, followed by Lys-D and Lys-L. These results are promising and indicate that the presented strategy is viable for designing and evaluating longer complementary peptides, as well as complementary peptides for co-assembly with other polypeptides of interest and importance. We discuss the possibility of designing complementary peptides to inhibit toxic amyloidosis with this approach.     

Modified Two-Phase Titration Methods to Quantify Surfactant Concentrations in Chemical-Enhanced Oil Recovery Applications

In enhanced oil recovery applications, surfactants are injected into reservoirs along with polymers and salts. The effluents eluted from lab experiments and field tests are analyzed by HPLC methods using an evaporative light scattering detector (ELSD) detector. When the surfactant concentrations are less than 100 ppm, HPLC methods are inaccurate. A novel two-phase titration method is developed where surfactant concentrations can be quantified using a calibration curve constructed with UV/vis absorption. This method can analyze surfactant concentrations 5–80 ppm where dilution eliminates any high-salinity interferences with the absorption measurements. The method is based on formation of a dye-surfactant complex and the light absorption of the complex has a linear correlation with the surfactant concentration. Anionic surfactant concentrations lower than 100 ppm can be accurately quantified using this method with methylene blue. The method was also developed for low concentrations (<50 ppm) of cationic surfactants using methyl orange and indigo carmine. The indigo carmine method can be used without the use of an organic phase. All methods are applicable at salinities up to 3 wt%. Both the methylene blue method and the methyl orange method can be used to detect zwitterionic surfactants. These methods can be used in the presence of polymers without any prior treatments.     

Metallic Nanoparticle-Enabled Sensing of a Drug-of-Abuse: An Attempt at Forensic Application

γ-Hydroxybutyric acid (GHB) functions as a depressant on the central nerve system and serves as a pharmaceutical agent in the treatment of narcolepsy and alcohol withdraw. In recent years, GHB has been misused as a recreational drug due to its ability to induce euphoric feelings. Moreover, it has gained increasing attention as a popular drug of abuse that is frequently related to drug-facilitated sexual assaults. At the moment, detection methods based on chromatography exhibit extraordinary sensitivity for GHB sensing. However, such techniques require complicated sample treatment prior to analysis. Optical sensors provide an alternative approach for rapid and simple analysis of GHB samples. Unfortunately, currently reported probes are mostly based on hydrogen bonding to recognize GHB, and this raises concerns about, for example, the lack of specificity. In this work, we report a bioinspired strategy for selective sensing of GHB. The method is based on specific enzyme recognition to allow highly selective detection of GHB with minimum interference, even in a complex sample matrix (e. g., simulated urine). In addition, the result can be obtained by either quantitative spectroscopy analysis or colorimetric change observed by the naked-eye, thus demonstrating its potential application in drug screening and forensic analysis.     

Process monitoring method based on correlation variable classification and vine copula

Chemical processes are becoming increasingly complicated, leading to an increase in process variables and more complex relationships among them. The vine copula has a significant advantage in portraying the dependence of high-dimensional variables. However, as the dimensions increase, the vine copula model incurs a high computational load; such pressure greatly reduces model efficiency. Relationships among variables in the industrial process are complex. Different variables may be strongly or weakly associated or even independent. This paper proposes a process monitoring method based on correlation variable classification and vine copula. The weighted correlation measure is first used to divide variables into a correlated subspace and weakly correlated subspace. Then, two vine structures, C-vine and D-vine, are applied to the correlated and weakly correlated subspaces, respectively. This method takes advantage of C-vine for correlated variables and the flexibility of D-vine for weakly correlated variables. Finally, comprehensive statistics are established based on different subspaces. Monitoring results of the numerical system and the Tennessee Eastman process demonstrate the effectiveness and validity of the proposed method.     

电磁精细探测法在隐伏型导水地质裂缝勘探中的应用

针对传统隐伏型导水地质裂缝勘探方法存在勘测精度较差的问题,提出电磁精细探测法探析隐伏型导水地质裂缝。依照屏蔽系数、实测场强和理论场强数据绘制综合曲线图,通过该图获取隐伏型导水地质裂缝所处位置几何阴影范围,采用层析成像法得到网格化的工作面,获取隐伏型导水地质图像。通过图像直接观测隐伏型导水地质工作面裂缝所处位置,在此基础上,观测四个电磁场分量,采用正交电磁场分量计算介质视电阻率,依据计算视电阻率数值和视电阻率分布状态研究裂缝发育情况和裂缝富水程度。结果表明:采用该方法能较为精准地获取隐伏型导水地质裂缝位置。通过裂缝位置进一步检测出隐伏型导水地质裂缝最大发育高度为63.5 m。当视电阻数值不断增加时,隐伏型导水地质裂缝和裂缝富水性逐渐减小,与实际情况较为相符,说明该种方法探析效果较好。     

基于量子线路逻辑运算的单光子量子态测量方案

在量子探测领域,关键任务之一就是要对未知量子态进行测量以获取量子态信息。通过将量子计算中的量子门所组成的量子线路应用于量子探测领域,提出实现单光子未知量子态的测量方案。利用量子计算的叠加性、纠缠性、可纠错性以及量子线路的可集成性,可以使得探测更具高效性并简化探测的实验系统。利用本文提出的探测新方案,通过仿真计算对该方案进行了模拟。在该方案的理论计算与仿真模拟结果的基础上,得到了以下的结论:通过在不同信噪比等参数的条件下选择适当的测量次数,基于量子线路的方案可以得到较为精确的测量结果。     

Proton-Detected Solid-State NMR of the Cell-Free Synthesized α-Helical Transmembrane Protein NS4B from Hepatitis C Virus

Proton-detected 100 kHz magic-angle-spinning (MAS) solid-state NMR is an emerging analysis method for proteins with only hundreds of microgram quantities, and thus allows structural investigation of eukaryotic membrane proteins. This is the case for the cell-free synthesized hepatitis C virus (HCV) nonstructural membrane protein 4B (NS4B). We demonstrate NS4B sample optimization using fast reconstitution schemes that enable lipid-environment screening directly by NMR. 2D spectra and relaxation properties guide the choice of the best sample preparation to record 2D ¹H-detected ¹H,¹⁵N and 3D ¹H,¹³C,¹⁵N correlation experiments with linewidths and sensitivity suitable to initiate sequential assignments. Amino-acid-selectively labeled NS4B can be readily obtained using cell-free synthesis, opening the door to combinatorial labeling approaches which should enable structural studies.     

浅谈城市废水水质检测方法与应用

国家与地区的经济发展水平一定离不开工业的发展,我国作为工业大国,工业发展对于国家整体经济有着不可取代的作用。然而随着工业的发展,全球性的环境污染和生态破坏越来越严重。特别是我国经济正处于快速发展阶段,城市化进程不断推进,城市污染情况日益严重,城市污水治理工作逐渐被大众所关注,逐渐开始进入公众可能给事业。我国二十世纪七十年代开始逐渐开始现代城市污水的治理工作。因此,对城市污水的水质监测工作成为了极为重要的一个环节。从城市污水检测方式这一角度进行研究,分析当前城市污水水质检验方法提升实际效果的应用模式,希望能够更好的处理我国城市日益严峻废水处理问题。