主成分分析-支持向量机用于肝病分类模型

将主成分分析(PCA)用于肝功能检测数据特征提取,然后用支持向量机(SVM)对乙肝、丙肝、肝硬化、正常人样本建立分类模型。采用高斯径向基函数(RBF)为核函数,调节核函数参数C及σ以建立最佳支持向量机模型。该模型对训练集的识别率为99.3%,对预测集的预测率为96.4%。结果表明:PCA-SVM法建立的肝病分类模型能较好的区分乙肝、丙肝、肝硬化及正常人,且分类效果优于传统支持向量机及人工神经网络(ANN)分类模型。     

支持矢量机和线性判别分析对细胞穿透肽的识别

选取25条CPP和16条非CPP作为训练集样本, 以61条CPP和21条非CPP为预测集样本. 利用氨基酸的z-Scale对肽链进行编码, 分别使用原始72个自交叉协方差变量和它们的主成分矢量进行线性判别(LDA)和支持矢量机(SVM)分类研究. 当采用LDA方法时, 对于训练集的预测以及它们的留一法交互检验, 均获得比较优越的结果, 但对预测集的预测总的识别率的最优结果仅为57.3%. 分别利用主成分和原始变量集作为SVM的输入建立的非线性识别模型, 对训练集的总识别率分别为85.4%和100%, 留一法交互检验的总识别率分别为80.5%和75.6%, 对预测集的最优总识别正确率为74.4%. 识别结果表明SVM能够比较好的提取原始变量间的细微模式变化, 对CPP总的识别结果优于LDA.     

机器学习方法用于建立乙酰胆碱酯酶抑制剂的分类模型

我们构建了表征乙酰胆碱酯酶抑制剂分子组成、电荷、拓扑、几何结构及物理化学性质等特征的1559个描述符,通过Fischer Score排序过滤和Monte Carlo模拟退火法相结合进行变量筛选得到37个描述符,然后分别用支持向量学习机(SVM)、人工神经网络(ANN)和k-近邻(k-NN)等机器学习方法建立了乙酰胆碱酯酶抑制剂的分类预测模型.对于训练集的515个样本,通过五重交叉验证,各机器学习方法对正样本,负样本和总样本的平均预测精度分别为87.3%-92.7%,67.0%-81.0%和79.4%-88.2%;通过y-scrambling方法验证SVM模型是否偶然相关,结果正样本,负样本和总样本的平均预测精度分别为72.7%-82.5%,41.0%-53.0%和62.1%-69.1%,明显低于实际所建模型的预测精度,表明所建模型不存在偶然相关;对172个没有参与建模的外部独立测试样本,各机器学习方法对正样本,负样本和总样本的预测精度分别为93.3%-100.0%,74.6%-89.6%和86.1%-95.9%.所建模型中,SVM模型预测精度最好,且明显高于其它文献报道结果.     

近红外光谱法同时测定通天口服液中天麻素和芍药苷的含量

本文应用近红外光谱结合偏最小二乘法建立了同时测定通天口服液中天麻素与芍药苷含量的方法。以高效液相色谱(HPLC)法测定通天口服液样品中天麻素和芍药苷的化学参考值,随机抽取60个样本作校正集,20个样本作预测集。用偏最小二乘法(PLS)将校正集样本的近红外光谱与相应样本的天麻素和芍药苷含量分别相关联建立模型。结果表明,天麻素和芍药苷校正模型的决定系数分别为96.28%、94.55%,模型的交叉验证均方差分别为0.0336、0.00908,预测集的决定系数分别为94.23%、92.86%,预测集均方差分别为0.0453、0.00839。同时还做了模型的精密度实验,该方法能用于大批量样品的快速分析。     

粒子群最小二乘支持向量机结合偏最小二乘法用于芝麻油质量的鉴别

结合粒子群最小二乘支持向量机(PSO-LSSVM)与偏最小二乘法(PLS)提出一种基于气相色谱技术的新方法,对芝麻油进行真伪鉴别,并对掺伪品中掺假比例进行定量分析。采用主成分分析法(PCA)对857个样本的脂肪酸色谱数据进行分析,优选主成分作为最小二乘支持向量机(LSSVM)的输入向量。利用粒子群算法(PSO)优化LSSVM,构建芝麻油掺伪鉴别的两级分类模型,同时运用PLS建立掺伪芝麻油中掺伪油脂的定量校正模型,两级分类模型的准确率分别达到了100%和98.7%,定量分析模型的平均预测标准偏差(RMSEP)为3.91%。结果表明,本方法的鉴别准确性和模型泛化能力均优于经典的BP神经网络和支持向量机(SVM),可用于食用油脂加工和流通环节的质量控制,为食用油质量的准确鉴定提供了一条有效途径。     

支持向量机和线性判别分析用于禽流感病毒编码蛋白识别

以支持向量机(SVM)和线性判别分析(LDA)对200条禽流感病毒、100条B型流感和100条C型流感病毒蛋白共400条为训练集样本,从表征序列的200个整体与局部变量中以逐步(stepwise)方法选取24个变量作为LDA模型的输入建立线性识别模型,病毒蛋白总识别率达99.8%,留一法交互检验总识别率为99.4%.从原始200变量中经主成分分析得16个主成分作为SVM的输入,以径向基核函数(RBF)SVM建立非线性识别模型,病毒蛋白总识别率为99.8%,留一法交互检验总识别率为99.2%.以100条禽流感、50条B型流感和50条C型流感病毒编码蛋白质共200条为测试集样本,得LDA模型,对其总识别正确率为95.4%,SVM模型对其总识别正确率为96.5%.识别结果表明,两个模型都可较好识别禽流感病毒蛋白,并且SVM对禽流感病毒蛋白的识别结果优于LDA.     

基于最小二乘支持向量机的油页岩含油率近红外光谱分析

为了提高油页岩含油率近红外光谱分析建模的预测精度和稳定性,开展了基于最小二乘支持向量机(LS-SVM)建模方法的对比研究.采用主成分-马氏距离(PCA-MD)和基于蒙特卡洛采样(MCS)2种方法进行了奇异样本的检测,采用径向基核函数的LS-SVM、偏最小二乘(PLS)和反向传播神经网络(BPANN)3种方法进行建模方法对比.结果表明,对于64个油页岩岩芯样本,与PCA-MD方法相比,采用MCS方法剔除奇异样本后所建PLS模型的预测精度提高了28%.对于MCS方法剔除奇异样本后的58个样品,采用KennardStone法划分了44个样品的校正集和14个样品的预测集,采用2阶导数和标准化预处理方法,建立了100个LS-SVM的校正模型,模型的预测决定系数R2平均值达到0.90以上,高于PLS和BPANN模型的对应值;且R2的变化量(0.02)小于BPANN模型的对应值(0.32).因此,MCS奇异样本检测结合LS-SVM方法可提高油页岩含油率样本建模的精度和稳定性.     

“动态”近红外光谱结合深度学习图像识别和迁移学习的模式识别方法研究

该文以山羊绒与山羊绒/羊毛混纺织物以及纯棉与丝光棉织物为研究对象,使用其"动态"光谱,扩大类间的光谱差异信息,通过融合其同步和异步二维相关光谱,用多张动态光谱构造一张能反映细节化学差异信息的"化学图像"。使用GoogLeNet深度神经网络图像识别模型结合迁移学习,建立了一种光谱分类的新方法。收集了234个织物样品,制备水含量分别为0、5.4%、11.2%和16.3%的样本,同时采集样品的漫反射近红外光谱。使用干基样品的多种预处理光谱,利用线性分类方法簇类独立软模式识别(SIMCA)和非线性方法支持向量机(SVM),共建立了16个分类模型。其中,山羊绒与山羊绒/羊毛混纺织物的SIMCA和SVM最优预测正确率分别为63.33%和70.09%,纯棉与丝光棉织物的分别为71.02%和72.51%,均不能实现有效分类。新方法对山羊绒与山羊绒/羊毛混纺织物的预测正确率为92.59%,纯棉与丝光棉织物的为94.74%,获得了有效分类。该文首次将图像分类方法用于光谱分类识别,开辟了一种新的研究途径。针对实际应用能收集到的样品属于小样本,不能满足深度学习需要大数据样本的问题,使用迁移学习方法使深度学习框架适应了光谱分类(小样本),为人工智能领域中先进的识别技术用于解决化学问题提供了一个成功示范。     

支持向量机回归法在近红外光谱测定植物纤维原料中甲氧基含量中的应用

以26个植物纤维原料为实验材料,由20个样品作校正样品,采用径向基核函数方法对纤维原料中甲氧基含量与纤维原料样品近红外光谱进行支持向量机(SVM)回归建模.以所建SVM回归模型对6个纤维原料样品中甲氧基含量进行预测,回归模型的预测结果与采用改良的维伯克法确定的甲氧基含量的相关系数为0.977,预测样本集的标准偏差为0.43.将SVM回归模型的预测效果与PLS回归模型的预测结果进行比较,所建近红外光谱测定植物纤维原料中甲氧基含量的SVM回归模型可用于实际植物纤维原料样品的定量分析,且具有较好的分析效果.     

近红外光谱结合CARS变量筛选方法用于液态奶中蛋白质与脂肪含量的测定

采用CARS(Competitive adaptive reweighted sampling)变量筛选方法建模,显著提高了液态奶中蛋白质与脂肪近红外模型的预测精度。用蒙特卡罗采样(Monte-Carlo sampling)方法先剔除奇异样本,再对光谱进行中心化与Karl Norris滤波降噪处理,通过CARS方法筛选出与样本性质密切相关的变量,建立预测蛋白质与脂肪含量的偏最小二乘法(PLS)校正模型,并与未选变量的PLS模型进行比较。以定标集相关系数(r2)及交互验证均方残差(RMSECV)和预测误差均方根(RMSEP)作为判定依据,确定了蛋白质与脂肪的最佳建模条件。蛋白质与脂肪校正模型的相关系数分别为0.975 0、0.995 1,RMSECV分别为0.194 8、0.136 3,RMSEP分别为0.113 3、0.140 1,预测结果优于未选变量的PLS模型及其他选变量方法,有效简化了模型,适于液态奶中脂肪和蛋白质的快速、无损检测。     

α-Amino acid derived enaminones and their application in the synthesis of N-protected methyl 5-substituted-4-hydroxypyrrole-3-carboxylates and other heterocycles

A new and simple synthesis of novel N-protected methyl 5-substituted-4-hydroxypyrrole-3-carboxylates, which exist in equilibrium with their 4-oxo tautomers, has been developed in two steps starting from N-protected α-amino acids. The key intermediates are enaminones, which can also be isolated, characterized, and used for the construction of other functionalized heterocycles, before they spontaneously decompose to pyrrole products. 4-Hydroxypyrroles are prone to partial aerial oxidation but can be efficiently alkylated or reduced to stable polysubstituted pyrrolidine derivatives.     

Specific intramolecular aromatic CH insertion of diazosulfonamides

The chemoselectivity in the intramolecular CH insertion of various diazosulfonamides has been experimentally studied. The results reveal that the aliphatic 1,4-, 1,5-, or 1,6-C(sp³)?H insertions of diazosulfonamides are not accessible, while the aromatic 1,5-C(sp²)?H insertion can be realized specifically by adjusting the diazo-adjacent group. In addition, the general chemoselectivities in the intramolecular CH insertions of diazosulfonyl compounds are summarized. Generally, diazosulfones undergo both aromatic 1,5-C(sp²)?H and aliphatic 1,5- and 1,6-C(sp³)?H insertions, while diazosulfonates undergo aliphatic 1,5- and 1,6-C(sp³)?H insertions. However, diazosulfonamides only undergo aromatic 1,5-C(sp²)?H insertion.     

N-Heterocyclic carbene-palladacyclic complexes: synthesis,characterization and their applications in the C-N coupling and α-arylation of ketones using aryl chlorides

N-Heterocyclic carbene-palladacyclic complexes 3 were successfully achieved in a one-pot procedure under mild conditions. The structure of 3a was unambiguously confirmed by X-ray single crystal diffraction and it was an active catalyst in the Buchwald-Hartwig amination and α-arylation of ketones even at very low catalyst loadings (0.01?mol%).     

Iodine-mediated oxidative Pictet-Spengler reaction using terminal alkyne as the 2-oxoaldehyde surrogate for the synthesis of 1-aroyl-β-carbolines and fused-nitrogen heterocycles

An efficient iodine-mediated oxidative Pictet-Spengler reaction in dimethyl sulphoxide (DMSO) using terminal alkynes as the 2-oxoaldehyde surrogate for the synthesis of aryl (9H-pyrido[3,4-b]indol-1-yl)methanones is described. The scope of the protocol includes the total synthesis of Fascaplysin, Eudistomins Y₁ and Y₂. The methodology is extended for preparing pyrrolo[1,2-a]-quinoxaline and indolo[1,5-a]quinoxaline derivatives. The utility of 1-aroyl-β-carbolines was demonstrated by performing palladium-catalyzed β-carboline directed ortho-C(sp2)-H functionalization of the phenyl ring with thiomethyl (SMe) group using DMSO as source and for accessing 4-aryl-canthin-6-ones.     

Facile synthesis of bicyclic 1-arylpyrazol-5-ones

In this Letter, we described a facile method for constructing fused bicyclic 1-arylpyrazol-5-one ring system. We employed various methylene-containing carboxylic acids as the substrates and proved that the pyrazolone ring closure requires activated methylene group in intermediate II. Accordingly, a series of structurally diversified, fused bicyclic 1-arylpyrazol-5-ones was prepared in moderate to high yields using the requisite substrates.     

Synthesis of pyrrolo[1,2-a]pyrazines and pyrazino[1,2-a]indoles by Curtius reaction in Morita-Baylis-Hillman derivatives

Synthesis of substituted pyrrolo[1,2-a]pyrazines and pyrazino[1,2-a]indoles from the Morita-Baylis-Hillman derivatives of acrylates via saponification followed by Curtius reaction is described.     

正丁胺对激光染料荧光谱影响的研究

用正丁胺作为碳源,采用射频辉光放电制备碳膜,选用激光染料R6G和聚乙二醇混合液作为蒸气源,采用单源热蒸发,在蒸发室与染料同时沉积得到混合膜,用拉曼光谱和红外光谱分析了碳膜的结构和键合方式,分析表明:碳膜中存在胺基团和氢原子.混合膜的荧光谱测量结果表明,认为正丁胺对染料荧光谱的影响是因为胺基和氢原子的存在.     

Cu0-catalysed 1,3-dipolar cycloadditions of α-amino acid derived N,N-cyclic azomethine imines to ynones

A series of 20 CuAIAC reactions between eight 4-acylamino substituted pyrazolidine-3-one-1-azomethine imines and four terminal ynones were performed using Cu⁰ as catalyst. The corresponding fluorescent cycloadducts were obtained in very high yields upon simple workup. Thus, Cu-metal turned out to be a better catalyst than Cu^I in terms of yield and ease of isolation. Availability of azomethine imines, mild reaction conditions, and simple workup enable a “click” access to libraries of densely substituted 2,3-dihydro-1H,5H-pyrazolo[1,2-a]pyrazol-1-ones. Reactivity of differently substituted dipoles was evaluated experimentally and by quantum chemical methods (DFT).     

Synthesis and antioxidative/anti-inflammatory activity of novel fullerene–polyamine conjugates

(E)-4-(Fullerenopyrrolidin-1-yl)-3-methylbut-2-enoic acid and its corresponding succinimidyl ester, readily obtained through Prato-type modification of C₆₀, were used for the selective N-acylation of polyamines. The thus obtained conjugates were evaluated for their antioxidative and anti-inflammatory activity and their cytotoxicity was determined. Members of this family of compounds showed interesting anti-lipid peroxidation, anti-lipoxygenase and anti-inflammatory activity and comparable cytocompatibility to spermidine.