基于快速神经网络架构搜索的鲁棒图像水印网络算法

为解决深度学习在图像水印算法中计算量大且模型冗余的问题，提高图像水印算法在抵抗噪声、旋转和剪裁等攻击时的鲁棒性，提出基于快速神经网络架构搜索（neural architecture search，NAS）的鲁棒图像水印网络算法。通过多项式分布学习快速神经网络架构搜索算法，在预设的搜索空间中搜索最优网络结构，进行图像水印的高效嵌入与鲁棒提取。首先，将子网络中线性连接的全卷积层设置为独立的神经单元结构，并参数化表示结构单元内节点的连接，预先设定结构单元内每个神经元操作的搜索空间；其次，在完成一个批次的数据集训练后，依据神经元操作中的被采样次数和平均损失函数值动态更新概率；最后，重新训练搜索完成的网络。水印网络模型的参数量较原始网络模型缩减了92%以上，大大缩短了模型训练时间。由于搜索得到的网络结构更为紧凑，本文算法具有较高的时间性能和较好的实验效果，在隐藏图像时，对空域信息的依赖比原始网络更少。对改进前后的2个网络进行了大量鲁棒性实验，对比发现，本文算法在CIFAR-10数据集上对抵抗椒盐噪声和旋转、移除像素行（列）等攻击优势显著；在ImageNet数据集上对抵抗椒盐高斯噪声、旋转、中值滤波、高斯滤波、JPEG压缩、裁剪等攻击优势显著，特别是对随机移除行（列）和椒盐噪声有较强的鲁棒性。     

On the solution of a mixed nonlinear integral equation

In this paper, we consider a mixed nonlinear integral equation of the second kind in position and time. The existence of a unique solution of this equation is discussed and proved. A numerical method is used to obtain a system of Harmmerstein integral equations of the second kind in position. Then the modified Toeplitz matrix method, as a numerical method, is used to obtain a nonlinear algebraic system. Many important theorems related to the existence and uniqueness solution to the produced nonlinear algebraic system are derived. The rate of convergence of the total error is discussed. Finally, numerical examples when the kernel of position takes a logarithmic and Carleman forms, are presented and the error estimate, in each case, is calculated.     

Multivariate control charts based on NAS and mid‐infrared spectroscopy for quality control of B5 blends of methyl soybean biodiesel in diesel

In this work, mid‐infrared spectroscopy and multivariate control charts based on net analyte signal were applied for quality control of B5 blends of biodiesel/diesel (5% biodiesel/95% diesel). Control charts were constructed using instrumental signal decomposition, generating three charts: the net analyte signal chart for monitoring the analyte of interest (methyl soybean biodiesel); the interference chart, which corresponds to the contribution of all other compounds in the diesel sample (diesel); and the residual chart, which corresponds to non‐systematic variations. Statistical limits were established for each developed chart, using samples inside quality specifications (normal operation conditions). To validate multivariate control charts, new samples were analyzed. The new samples represented samples in‐control and samples out‐of‐control in relation to the content of biodiesel, adulterated biodiesel with severe vegetable oils and adulterated diesel with residual automotive lubricant oil, kerosene, and gasoline. The results obtained show an excellent distinction between the samples inside and out of the quality specifications, with 91% and 100% correctly classified, respectively, which demonstrates that the methodology developed is a viable alternative for quality monitoring of this type of fuel. Copyright © 2015 John Wiley & Sons, Ltd.     

Improving the robustness of a partial least squares (PLS) model based on pure component selectivity analysis and range optimization: case study for the analysis of an etching solution containing hydrogen peroxide

Pure component selectivity analysis (PCSA) was successfully utilized to enhance the robustness of a partial least squares (PLS) model by examining the selectivity of a given component to other components. The samples used in this study were composed of NH₄OH, H₂O₂ and H₂O, a popular etchant solution in the electronic industry. Corresponding near-infrared (NIR) spectra (9000-7500 cm⁻¹) were used to build PLS models. The selective determination of H₂O₂ without influences from NH₄OH and H₂O was a key issue since its molecular structure is similar to that of H₂O and NH₄OH also has a hydroxyl functional group. The best spectral ranges for the determination of NH₄OH and H₂O₂ were found with the use of moving window PLS (MW-PLS) and corresponding selectivity was examined by pure component selectivity analysis. The PLS calibration for NH₄OH was free from interferences from the other components due to the presence of its unique NH absorption bands. Since the spectral variation from H₂O₂ was broadly overlapping and much less distinct than that from NH₄OH, the selectivity and prediction performance for the H₂O₂ calibration were sensitively varied depending on the spectral ranges and number of factors used. PCSA, based on the comparison between regression vectors from PLS and the net analyte signal (NAS), was an effective method to prevent over-fitting of the H₂O₂ calibration. A robust H₂O₂ calibration model with minimal interferences from other components was developed. PCSA should be included as a standard method in PLS calibrations where prediction error only is the usual measure of performance.     

Radiation grafting of NIPAAm and acryloxysuccinimide onto PP films and sequent crosslinking with polylysine

N-isopropylacrylamide (NIPAAm) and N-acryloxysuccinimide (NAS) were grafted from their binary mixtures in tetrahydrofurane (THF) and toluene solutions onto polypropylene (PP) films by the pre-irradiation oxidative method in air. Effects of pre-irradiation dose, dose rate, and monomer concentrations (NAS/NIPAAm) were studied. The grafted copolymers exhibited the lower critical solution temperature (LCST) at around 31 °C. Based on its thermo-reversible behavior, this system has been used for immunoassay, drug delivery, separation processes and immobilization of enzymes. N-acryloxysuccinimide (NAS) has been used as an active ester to bind proteins through amide bond formation with lysine, and because of this property, the grafted copolymer has been crosslinked with polylysine. Techniques used to characterize the films included differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), infrared (FTIR-ATR) and elemental analysis. Results on thermo-sensitivity are presented. This new system could find applications in vesicle immobilizations.     

Finding Y‐relevant part of X by use of PCR and PLSR model reduction methods

The paper is considering the following question: using principal component regression (PCR) or partial least squares regression (PLSR), how much data can be removed from X while retaining the original ability to predict Y ? Two model reduction methods using similarity transformations are discussed, one giving projections of original loadings onto the column space of the fitted response matrix (essentially the orthogonal signal correction (OSC) methods), and one giving projections of original scores onto the column space of the coefficient matrix (essentially the net analyte signal (NAS) methods). The loading projection method gives model residuals that are orthogonal to Y and , which is valuable in certain applications. The score projection method, on the other hand, gives model residuals that are orthogonal to , which is essential in other applications. It is shown that the reduced matrix X from the score projection method is a subset of the reduced matrix X from the loading projection method. It therefore has the smallest Frobenius norm, and thus the smallest total column variance, assuming centered data. Copyright © 2007 John Wiley & Sons, Ltd.     

Synthesis and characterization of thermosensitive interpenetrating polymer networks based on N-isopropylacrylamide/N-acryloxysuccinimide, crosslinked with polylysine, grafted onto polypropylene

Interpenetrating polymer networks (IPNs) based on poly (N-isopropylacrylamide), (PNIPAAm) and poly (N-acryloxysuccinimide) (PNAS), grafted onto polypropylene (PP), were synthesized in three consecutive steps using ionizing radiation in the first and second steps and chemical reaction in third one. In the first step a thermosensitive graft copolymer of NIPAAm onto PP film was obtained by gamma radiation with a ⁶⁰Co source. The grafted side chains of PNIPAAm were then crosslinked with gamma radiation to give net-[PP-g-NIPAAm]. The secondary network was obtained in situ by chemical crosslinking between PNAS and polylysine (pLys). The PP-g-IPNs exhibited the lower critical solution temperature (LCST) at around 32 °C. Based on its thermoreversible behavior, this system could be used for immobilization of biomolecules. The phase transition temperature (LCST) and network properties of the IPNs were measured by swelling behavior. Additional characterization by differential scanning calorimetry (DSC), thermogravimetric analysis (TGA) and infrared (FTIR-ATR) determinations are reported.     

Interrelationships between generalized Tikhonov regularization,generalized net analyte signal,and generalized least squares for desensitizing a multivariate calibration to interferences

Orthogonal pre‐processing (orthogonal projection) of spectral data is a common approach to generate analyte‐specific information for use in multivariate calibration. The goal of this pre‐processing is to remove from each spectrum the respective sample interferent contributions (spectral interferences from overlap, scatter, noise, etc.). Two approaches to accomplish orthogonal pre‐processing are net analyte signal (NAS) and generalized least squares (GLS). Developed in this paper is the mathematical relationship between NAS and GLS. It is also realized that orthogonal NAS pre‐processing can remove too much analyte signal and that the degree of interferent correction can be regulated. Similar to GLS, the degree of correction is accomplished by using a regularization (tuning) parameter to form generalized NAS (GNAS). Also developed in this paper is an alternative to GNAS and GLS based on generalized Tikhonov regularization (GTR). The mathematical relationships between GTR, GNAS, and GLS are derived. A result is the ability to express the model vector as the sum of two contributions: the orthogonal NAS contribution and a non‐NAS contribution from the interferent components. Thus, rather than the usual situation of sequentially pre‐processing data by either GNAS or GLS followed by model building with the pre‐processed data, the methods of GTR, GNAS, and GLS are expressed as direct computations of model vectors allowing concurrent pre‐processing and model building to occur. Simultaneous pre‐processing and model forming are shown to be natural to the GTR process. Two near‐infrared spectroscopic data sets are studied to compare the theoretical relationships between GTR, GNAS, and GLS. One data set covers basic calibration, and the other data set is for calibration maintenance. Filter factor representation is key to developing the interprocess relationships. Copyright © 2013 John Wiley & Sons, Ltd.     

UHPLC‐MS/MS quantification of buprenorphine,norbuprenorphine, methadone,and glucuronide conjugates in umbilical cord plasma

Opioid use during pregnancy can result in the newborn being physically dependent on the substance, thus experiencing drug withdrawal, termed neonatal abstinence syndrome (NAS). Buprenorphine and methadone are two drugs used to treat opioid withdrawal and are approved for use in pregnancy. Quantification of these compounds in umbilical cord plasma would help assess in utero exposure of neonates in cases of buprenorphine or methadone use during pregnancy. An LC‐MS/MS method using solid‐phase extraction sample preparation was developed and validated for the simultaneous quantification of methadone, buprenorphine, norbuprenorphine, and glucuronide metabolites in umbilical cord plasma. The average accuracy (percentage error) and precision (relative standard deviation) were <15% for each validated concentration. Our data establishes a 2 week maximum freezer storage window in order to achieve the most accurate cord plasma concentrations of these analytes. Additionally, we found that the umbilical cord tissue analysis was less sensitive compared with analysis with umbilical cord blood plasma, indicating that this may be a more appropriate matrix for determination of buprenorphine and metabolite concentrations. This method was successfully applied to the analysis of cord blood from women with known buprenorphine or methadone use during pregnancy. Copyright © 2015 John Wiley & Sons, Ltd.     

近红外分析中光谱预处理及波长选择方法进展与应用

光谱预处理和波长选取方法在近红外光谱分析技术中相当重要。本文综述了常用的NIR预处理和波长选取方法及这一领域的最新进展，详细介绍正交信号校正(OSC)、净分析信号(NAS)和小波变换(WT)等新光谱预处理方法以及无信息变量消除(UVE)和遗传算法(GA)等波长选取方法，并给出了这些方法的具体算法和一些应用实例。