Quantitative structure-property relationship study of the solubility of thiazolidine-4-carboxylic acid derivatives using ab initio and genetic algorithm-partial least squares

A quantitative structure-activity relationships （QSAR） study is suggested for the prediction of solubility of some thiazolidine-4- carboxylic acid derivatives in aqueous solution. Ab initio theory was used to calculate some quantum chemical descriptors including electrostatic potentials and local charges at each atom, HOMO and LUMO energies, etc. Modeling of the solubility of thiazolidine- 4-carboxylic acid derivatives as a function of molecular structures was established by means of the partial least squares （PLS）. The subset of descriptors, which resulted in the low prediction error, was selected by genetic algorithm. This model was applied for the prediction of the solubility of some thiazolidine-4-carboxylic acid derivatives, which were not in the modeling procedure. The relative errors of prediction lower that -4% was obtained by using GA-PLS method. The resulted model showed high prediction ability with RMSEP of 3.8836 and 2.9500 for PLS and GA-PLS models, respectively.

基于MCC-GAPLS-PLSR的辣椒叶绿素含量高光谱定量反演

为了准确监测辣椒生长,本研究对辣椒冠层光谱反射率进行对数处理、倒数处理、倒数的对数处理、连续统去除处理、一阶微分处理、二阶微分处理,并与SPAD值进行相关性分析,用最大相关系数法(MCC)选取相关性较好的特征波段生成特征波段数据集,再用遗传算法-偏最小二乘法(GAPLS)进行降维得到最优特征波段组合,采用偏最小二乘法(PLSR)、反向传播神经网络(BPNN)、随机森林(RF)和最小二乘支持向量机(LSSVM)4种机器学习算法构建辣椒叶绿素含量反演模型。结果表明,最优波段和对应处理分别为700 nm(原始光谱)、699 nm(对数处理)、713 nm(连续统去除处理)、500 nm(二阶微分处理)、713 nm(二阶微分处理)。GAPLS的降维效果较好,与降维前相比PLSR模型的精度提升率最高,R^(2)、RPD分别提升了82.22%、136.98%,RMSE降低了29.96%。4种模型中,GAPLS降维处理后的PLSR模型的精度最好,R^(2)、RMSE和RPD分别为0.82、1.94、4.55。本研究构建的MCC-GAPLS-PLSR模型具有较好的反演潜力,适用于研究区辣椒叶片叶绿素含量测定,推动辣椒高效种植。

Factors Affecting Box Office during Broad Spring Festival Based on Partial Least Squares Regression

The box office during the later Spring Festival shows an attractive prospect.This paper studied the factors affecting total box office during the broad Spring Festival which is from the Spring Festival to the Lantern Festival.Data of films released during the broad Spring Festival from the years 2016 to 2019 in China were gathered,and the impact of eight explanatory variables on the box office during the broad Spring Festival was empirically analyzed by partial least squares(PLS)regression with software SIMCA.The results suggest that word-of-mouth has the most positive effect on the box office during the broad Spring Festival.Later propaganda has a positive effect,while early promotion has a negative effect on the box office.Director’s influence has a positive effect,while actor’s influence does not contribute much to the box office.Length of the trailer has a negative effect.The film format of 2D or 3D doesn’t contribute much to the box office.

Boosting the partial least square algorithm for regression modelling

Boosting algorithms are a class of general methods used to improve the general periormance of regression analysis. The main idea is to maintain a distribution over the train set. In order to use the given distribution directly, a modified PLS algorithm is proposed and used as the base learner to deal with the nonlinear multivariate regression problems. Experiments on gasoline octane number prediction demonstrate that boosting the modified PLS algorithm has better general performance over the PLS algorithm.

基于MW-MKEPLS的多重时变间歇生产过程质量预测

间歇生产过程的多重时变特性和非线性使得质量预测问题变得复杂。为了提高间歇过程质量预测精度,提出了滑动窗多向核熵偏最小二乘(moving window multiway kernel entropy partial least squares,MW-MKEPLS)方法。首先采用滑动窗进行数据的动态更新获取,构建了滑动窗多重时变模型;然后在滑动窗多重时变模型下通过核函数将数据映射到高维特征空间,采用Renyi熵贡献度进行数据特征提取,更好地获取数据的信息熵和非线性;最后在KECA处理后的高维特征空间进行质量预测。通过青霉素生产发酵过程进行了实验验证,并与MKPLS和MKEPLS进行对比分析,结果表明所提方法的质量预测精度更高。

基于microRNA表达谱初步构建PLS-DA体液识别模型

针对外周血、唾液、精液、月经血、阴道分泌物这五种法医学常见的体液样本,采用二代测序(NGS)平台检测获得microRNA(miRNA)表达谱,使用偏最小二乘判别分析(PLS-DA)建立基于这五种体液的识别模型,并探讨PLS-DA在法医体液溯源中的应用价值。经优化小分子RNA文库制备过程,采用Ion Torrent S5 XL测序系统对前述法医学五种体液样本(每种10例)进行小RNA测序,以PLS-DA构建体液识别模型,评估不同数量miRNA标记组合下预测的准确性。本研究获得法医学常见五种体液样本的miRNA表达谱,外周血与月经血中表达量前10名的miRNAs有6个重叠;唾液和阴道分泌物中表达量前10名的miRNAs有4个重叠。基于全数据集、107个和11个miRNAs构建的体液来源识别模型的准确率分别为0.95、0.94、0.89。本研究通过NGS测序分析获得了五种体液样本的miRNA组(miRNome),利用PLS-DA初步构建了体液识别模型,对于应用miRNome进行体液识别的相关研究具有参考价值。

Characterizing and estimating rice brown spot disease severity using stepwise regression,principal component regression and partial least-square regression

Detecting plant health conditions plays a key role in farm pest management and crop protection. In this study, measurement of hyperspectral leaf reflectance in rice crop (Oryzasativa L.) was conducted on groups of healthy and infected leaves by the fungus Bipolaris oryzae (Helminthosporium oryzae Breda. de Hann) through the wavelength range from 350 to 2 500 nm. The percentage of leaf surface lesions was estimated and defined as the disease severity. Statistical methods like multiple stepwise regression, principal component analysis and partial least-square regression were utilized to calculate and estimate the disease severity of rice brown spot at the leaf level. Our results revealed that multiple stepwise linear regressions could efficiently estimate disease severity with three wavebands in seven steps. The root mean square errors (RMSEs) for training (n=210) and testing (n=53) dataset were 6.5% and 5.8%, respectively. Principal component analysis showed that the first principal component could explain approximately 80% of the variance of the original hyperspectral reflectance. The regression model with the first two principal components predicted a disease severity with RMSEs of 16.3% and 13.9% for the training and testing dataset, respec-tively. Partial least-square regression with seven extracted factors could most effectively predict disease severity compared with other statistical methods with RMSEs of 4.1% and 2.0% for the training and testing dataset, respectively. Our research demon-strates that it is feasible to estimate the disease severity of rice brown spot using hyperspectral reflectance data at the leaf level.

Correlation analysis and partial least square modeling to quantify typical minerals with Chang'E-3 visible and near-infrared imaging spectrometer's ground validation data

In 2013, Chang'E-3 program will develop lunar mineral resources in-situ detection. A Visible and Near-infrared Imaging Spectrometer(VNIS) has been selected as one payload of CE-3 lunar rover to achieve this goal. It is critical and urgent to evaluate VNIS' spectrum data quality and validate quantification methods for mineral composition before its launch. Ground validation experiment of VNIS was carried out to complete the two goals, by simulating CE-3 lunar rover's detection environment on lunar surface in the laboratory. Based on the hyperspectral reflectance data derived, Correlation Analysis and Partial Least Square(CA-PLS) algorithm is applied to predict abundance of four lunar typical minerals(pyroxene, plagioclase, ilmenite and olivine) in their mixture. We firstly selected a set of VNIS' spectral parameters which highly correlated with minerals' abundance by correlation analysis(CA), and then stepwise regression method was used to find out spectral parameters which make the largest contributions to the mineral contents. At last, functions were derived to link minerals' abundance and spectral parameters by partial least square(PLS) algorithm. Not considering the effect of maturity, agglutinate and Fe0, we found that there are wonderful correlations between these four minerals and VNIS' spectral parameters, e.g. the abundance of pyroxene correlates positively with the mixture's absorption depth, the value of absorption depth added as the increasing of pyroxene's abundance. But the abundance of plagioclase correlates negatively with the spectral parameters of band ratio, the value of band ratio would decrease when the abundance of plagioclase increased. Similar to plagioclase, the abundance of ilmenite and olivine has a negative correlation with the mixture's reflectance data, if the abundance of ilmenite or olivine increase, the reflectance values of the mixture will decrease. Through model validation, better estimates of pyroxene, plagioclase and ilmenite's abundances are given. It is concluded that VNIS has the capability to be applied on lunar minerals' identification, and CA-PLS algorithm has the potential to be used on lunar surface's in-situ detection for minerals' abundance prediction.

Utilizing partial least square and support vector machine for TBM penetration rate prediction in hard rock conditions

Rate of penetration(ROP) of a tunnel boring machine(TBM) in a rock environment is generally a key parameter for the successful accomplishment of a tunneling project. The objectives of this work are to compare the accuracy of prediction models employing partial least squares(PLS) regression and support vector machine(SVM) regression technique for modeling the penetration rate of TBM. To develop the proposed models, the database that is composed of intact rock properties including uniaxial compressive strength(UCS), Brazilian tensile strength(BTS), and peak slope index(PSI), and also rock mass properties including distance between planes of weakness(DPW) and the alpha angle(α) are input as dependent variables and the measured ROP is chosen as an independent variable. Two hundred sets of data are collected from Queens Water Tunnel and Karaj-Tehran water transfer tunnel TBM project. The accuracy of the prediction models is measured by the coefficient of determination(R2) and root mean squares error(RMSE) between predicted and observed yield employing 10-fold cross-validation schemes. The R2 and RMSE of prediction are 0.8183 and 0.1807 for SVMR method, and 0.9999 and 0.0011 for PLS method, respectively. Comparison between the values of statistical parameters reveals the superiority of the PLSR model over SVMR one.

Comparison of Calibration Curve Method and Partial Least Square Method in the Laser Induced Breakdown Spectroscopy Quantitative Analysis

The Laser Induced Breakdown Spectroscopy (LIBS) is a fast, non-contact, no sample preparation analytic technology;it is very suitable for on-line analysis of alloy composition. In the copper smelting industry, analysis and control of the copper alloy concentration affect the quality of the products greatly, so LIBS is an efficient quantitative analysis tech- nology in the copper smelting industry. But for the lead brass, the components of Pb, Al and Ni elements are very low and the atomic emission lines are easily submerged under copper complex characteristic spectral lines because of the matrix effects. So it is difficult to get the online quantitative result of these important elements. In this paper, both the partial least squares (PLS) method and the calibration curve (CC) method are used to quantitatively analyze the laser induced breakdown spectroscopy data which is obtained from the standard lead brass alloy samples. Both the major and trace elements were quantitatively analyzed. By comparing the two results of the different calibration method, some useful results were obtained: both for major and trace elements, the PLS method was better than the CC method in quantitative analysis. And the regression coefficient of PLS method is compared with the original spectral data with background interference to explain the advantage of the PLS method in the LIBS quantitative analysis. Results proved that the PLS method used in laser induced breakdown spectroscopy was suitable for simultaneous quantitative analysis of different content elements in copper smelting industry.

基于IPLS-XGBoost算法的可见-近红外光谱鸡蛋新鲜度高效准确检测技术研究

针对传统光谱法检测鸡蛋新鲜度存在的效率低、准确率不够高等问题,提出采用可见-近红外光谱结合极度提升树(XGBoost)等算法对鸡蛋新鲜度分类进行研究,以期在保证足够高准确度的同时大幅提高检测效率。将不同储存条件下的鸡蛋作为样本,并分别划分为训练集和测试集,采用训练集的综合评价指标(F-measure)和准确率(Accuracy)评估分类模型的性能。具体地,首先利用可见-近红外光谱系统采集鸡蛋的反射光谱,将所得的光谱数据经过不同预处理后再结合随机森林(random forest,RF)、偏最小二乘(partial least squares,PLS)、支持向量机(support vector machine,SVM)、多层感知机(muhi-layer perception,MLP)以及XGBoost等分类算法构建鸡蛋新鲜度分类评估模型,并对比各模型性能指标。分析结果发现,经Savitzky-Golay一阶导(Savitzky Golay first-order derivative,SG-1^(st)-Der)预处理后的RF、SVM、XGBoost模型和经标准正态变量(standardized normal variate,SNV)预处理后的PLS、MLP模型具有较好的训练结果。为进一步提高模型精度和运算效率,提出利用区间偏最小二乘法(interval partial least squares,IPLS)对SG-1^(st)-Der和SNV预处理后的光谱数据首先进行降维,然后再分别建立基于RF、SVM、XGBoost、PLS及MLP等算法的预估模型,最后通过测试集对模型进行验证。结果发现原始光谱数据经SG-1^(st)-Der预处理后所建立的IPLS-XGBoost分类模型性能最优,在不同储藏条件下测试集的F-measure分别为92.33%和90%,Accuracy分别达到94.44%和91.67%,而程序运行时间均不超过0.6 s。表明,可见-近红外光谱结合IPLS-XGBoost分类算法可应用于鸡蛋新鲜度评估,该方法在模型分类性能、准确度评估、运行速度等方面比传统方法更具优越性。

近红外光谱技术结合iPLS_SPA波段筛选在黄水酒精度预测模型中的应用

为实现白酒发酵过程中黄水酒精度的快速检测,研究采用傅里叶近红外光谱(FT-NIR)技术对黄水进行光谱采集,并且采用偏最小二乘回归(PLSR)法建立酒精度预测模型。为减少全光谱的数据冗余降低复杂度,提升建模准确率,将连续投影算法(SPA)与间隔偏最小二乘法(iPLS)联用,对整个谱区进行特征波段筛选,并用决定系数R2与预测均方根误差(RMSEP)评价预测模型。结果表明:与原始数据集相比,经过异常样品剔除、预处理、特征光谱筛选后预测模型,预测集R2也从最开始的0.702变为0.952,提升35.61%;预测RMSEP从3.812变为1.367,降低64.14%;变量数也从2,203逐步下降到99,降低了95.51%。说明在减少非相关信息与噪声的同时,模型的复杂度也得到极大改善,并且模型的稳定性与准确度得到了有效提升,最终实现黄水酒精度的快速无损检测,以期为白酒发酵领域提供一种新的可能性,为近红外在白酒发酵副产物中的检验提供理论基础。

滇中高原湖泊湖滨带土壤生态化学计量特征及影响因素研究

湖滨带是地球化学元素循环高度活跃且极易受环境变化与人类活动影响的敏感地带。文章通过对滇中湖泊杞麓湖湖滨带土壤碳(C)、氮(N)、磷(P)含量的分析,结合冗余分析和结构方程模型等,研究了土壤生态化学计量的时空变化特征及其与土壤理化性质间的耦合关系。结果表明,杞麓湖湖滨带土壤总有机碳(TOC)、总氮(TN)、总磷(TP)平均含量分别为31.97、1.79、1.01 g/kg。雨季土壤TOC、TN含量低于旱季,在湖滨带结构上均呈现水向带>消落带>陆向带的特征,且随土壤深度的增加而降低,TP时空差异不显著。土壤C∶N、C∶P和N∶P平均值分别为23.08、41.38、2.10,表明湖滨带土壤有机质的分解速率较慢,释放有效磷的潜力大,植物生长易受到氮的限制,而磷的高含量将影响杞麓湖流域生态环境保护。土壤容重、含水率和黏土含量通过改变其C、N、P含量,进而影响土壤生态化学计量比。

偏最小二乘法(PLS)及其在分析化学中的应用

本文介绍了一种性能优于其他多元统计方法的新的多元校准法——偏最小二乘法(PLS),研究了其理论基础、算法和性能,评述了其在吸收光谱分析、荧光分光光度分析、ICP—AES、色谱分析、核磁共振谱分析、生物化学、产品质量预测、毒理学、环境化学和地球化学研究等方面的应用以及应用发展动向。引用文献58篇。

PLS-BP法近红外光谱定量分析研究

建立BP模型用于近红外光谱定量分析时,为克服所建模型与训练样本集产生“过拟合”,先用线性算法为其压缩训练数据是必要的。目前多采用主成分法(PCA)和逐步回归法(SRA)。主成分法具有极强的压缩数据能力,用它压缩成的主成分输入BP网所建模型的预测精度一般能满足要求,但它处理数据时未考虑输出变量的影响。逐步回归法根据系统输出选择变量,但所选变量具有自相关性,而且与训练集样品的排列顺序有关,很难选出最好的变量,往往难满足预测精度要求。本研究用偏最小二乘法(PLS),根据输出变量将原始数据压缩为主成分,输入BP网并用所建模型预测30个小麦样品的蛋白质含量。结果表明,与PCA-BP模型的预测决定系数(R2)从92.50提高到97.10,训练迭代次数从12 000减少到4 500。

基于PLS算法的生物质秸秆元素分析NIRS快速检测

为探讨利用近红外光谱技术快速检测生物质秸秆中N、C、H、S和O 元素的可行性,采集并制备水稻、小麦、油菜和玉米秸秆样本199个,采用近红外光谱（NIRS）分析技术,结合偏最小二乘（PLS）化学计量学算法,在7400-5550cm^-1波段范围内,比较不同光谱预处理方法的定标效果,建立最优的生物质秸秆中N、C、H、S和O 元素的定量分析模型,并用独立的验证集样本对模型进行验证.验证结果表明所建立的N元素的定量分析模型可用于实际检测;O 元素的定量分析模型可进行实际估测;采用近红外技术用于C元素定量分析是可行的,但模型需要进一步优化;H、S元素采用NIRS技术无法进行定量分析.

PLS回归法分析多因素对卷烟燃烧温度及主流烟气有害成分释放量的影响

为了考察卷烟材料(卷烟纸、接装纸、成型纸和滤棒)和膨胀梗丝多因素作用对卷烟燃烧锥最高温度以及焦油、烟碱和主流烟气中7种有害成分(CO,HCN,NNK,NH3,B[a]P,苯酚和巴豆醛)释放量的影响,采用偏最小二乘(PLS)回归法建立了卷烟燃烧锥最高温度以及焦油、烟碱和主流烟气中7种有害成分释放量的多因素预测模型。结果表明:1卷烟燃烧锥最高温度的主要影响因素为卷烟纸助燃剂含量。随卷烟纸助燃剂含量增加,卷烟燃烧锥最高温度有降低趋势。2焦油、烟碱和主流烟气7种有害成分释放量的主要影响因素为接装纸透气度。随接装纸透气度增加,焦油、烟碱和主流烟气7种有害成分释放量有降低趋势。3随卷烟纸助燃剂和膨胀梗丝含量的增加,焦油、烟碱及主流烟气7种有害成分的释放量有降低趋势。适当增加接装纸透气度、卷烟纸助燃剂含量和膨胀梗丝含量,可在一定程度上调控卷烟燃烧锥温度,降低卷烟烟气有害成分释放量。

PLS-DA法判别分析木材生物腐朽的研究

利用近红外光谱结合PLS-DA判别分析方法可用于食品、药品和农产品等的快速识别或检测,因此,研究利用近红外光谱结合PLS-DA方法来检测木材的生物腐朽。研究结果表明:应用近红外光谱结合PLS-DA方法对培训集样本建立的判别模型,其校正及验证结果与实际分类变量的相关系数均超过0.94,SEC和SEP都低于0.17;利用模型对未参与建模的样本进行检测,发现该模型对未腐朽、白腐和褐腐三种类型样本的判别准确率均为100%(偏差均小于0.5);与SIMCA法相比,PLS-DA法对木材生物腐朽样本的判别准确率更高,说明应用近红外光谱结合PLS-DA方法能快速地检测到木材的生物腐朽,并能准确地判别出木材的生物腐朽类型。

IPLS-SPA波长选择方法在近红外秸秆生物量中的应用研究

在近红外光谱分析模型中全谱信息通常含有大量冗余信息,会导致模型解析时间延长、加大模型解析难度,因此如何快速有效地选取特征波长至关重要。采用基于间隔偏最小二乘（interval partial least squares,IPLS）结合连续投影算法（successive projections algorithm,SPA）对小麦秸秆发酵过程微生物生物量进行特征波长选择,共制备85个样本,采用氨基葡萄糖法测定微生物生物量,选择68个样本作为校正集,17个样本作为验证集。首先对全谱区520个波长点根据间隔点大小10,20,30,40进行分段建模,选取出4 450～4 925和9 194～9 993cm^-1两个波段范围作为特征波段,将选取出的特征波段再进行连续投影算法及遗传算法（genetic algorithm,GA）特征波长点选取,并进行综合分析对比。实验结果表明采用IPLS-SPA算法选择4 450～4 925和9 194～9 993cm^-1的组合波段具有最佳建模效果,相比于全谱建模其参与建模的波长点由520个减少到10个,模型验证集决定系数（R-Square,R2）从0.884 9提升至0.945 28,验证集均方误差根（root mean square error prediction,RMSEP）从11.104 9降至8.203 3,GA遗传算法虽取得了更优的模型精度,但其实验结果并不稳定且随机性较强,而IPLS结合SPA方法能够稳定而准确的（地）选择特征波长信息,提高模型运算速度并降低模型拟合难度,可以作为一种新的波段选择参考方法。结果表明采用近红外光谱分析方法对秸秆发酵生物量进行快速检测是可行的。

基于香溪河中游水体总磷总氮高光谱估算模型比较

研究通过地面高光谱遥感技术,针对香溪河中游的水质参数——总氮(TN)和总磷(TP)进行反演分析。研究选取160个数据样本,采用了单波段分析、一阶微分波段分析、波段比值分析、双波长差异指数、归一化差异指数以及偏最小二乘回归(PLS)分析6种方法建立反演模型。结果显示,单波段分析和一阶微分波段分析对于TP和TN的反演效果不佳。而波段比值分析在一定程度上提高了TN的反演准确性,其二次幂回归分析的R^(2)可达0.3674。在双波长差异指数的3次幂分析中,TP的预测模型达到最高R^(2)值,为0.4014。PLS回归分析在TN反演模型中表现突出,波段差值混合模型的R^(2)达到0.39,RMSEP为0.501,ARE为40.278%,展现一定的预测准确性。研究结果有助于利用高光谱遥感预测香溪河中游TN、TP长时间变化趋势,但模型预测精度仍受限于数据集和水体类型,未来研究需进一步探索和优化模型算法。