A partial least-squares regression approach to land use studies in the Suzhou-Wuxi-Changzhou region

In several LUCC studies, statistical methods are being used to analyze land use data. A problem using conventional statistical methods in land use analysis is that these methods assume the data to be statistically independent. But in fact, they have the tendency to be dependent, a phenomenon known as multicollinearity, especially in the cases of few observations. In this paper, a Partial Least-Squares （PLS） regression approach is developed to study relationships between land use and its influencing factors through a case study of the Suzhou-Wuxi-Changzhou region in China. Multicollinearity exists in the dataset and the number of variables is high compared to the number of observations. Four PLS factors are selected through a preliminary analysis. The correlation analyses between land use and influencing factors demonstrate the land use character of rural industrialization and urbanization in the Suzhou-Wuxi-Changzhou region, meanwhile illustrate that the first PLS factor has enough ability to best describe land use patterns quantitatively, and most of the statistical relations derived from it accord with the fact. By the decreasing capacity of the PLS factors, the reliability of model outcome decreases correspondingly.

PARTIAL LEAST-SQUARES(PLS)REGRESSION AND SPECTROPHOTOMETRY AS APPLIED TO THE ANALYSIS OF MULTICOMPONENT MIXTURES

The UV absorption spectra of o-naphthol,α-naphthylamine,2,7-dihydroxy naphthalene,2,4-dimethoxy ben- zaldehyde and methyl salicylate,overlap severely;therefore it is impossible to determine them in mixtures by traditional spectrophotometric methods.In this paper,the partial least-squares(PLS)regression is applied to the simultaneous determination of these compounds in mixtures by UV spectrophtometry without any pretreatment of the samples.Ten synthetic mixture samples are analyzed by the proposed method.The mean recoveries are 99.4%,996%,100.2%,99.3% and 99.1%,and the relative standard deviations(RSD) are 1.87%,1.98%,1.94%,0.960% and 0.672%,respectively.

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.

Partial least squares regression for predicting economic loss of vegetables caused by acid rain

To predict the economic loss of crops caused by acid rain,we used partial least squares(PLS) regression to build a model of single dependent variable -the economic loss calculated with the decrease in yield related to the pH value and levels of Ca2+,NH4+,Na+,K+,Mg2+,SO42-,NO3-,and Cl-in acid rain. We selected vegetables which were sensitive to acid rain as the sample crops,and collected 12 groups of data,of which 8 groups were used for modeling and 4 groups for testing. Using the cross validation method to evaluate the performace of this prediction model indicates that the optimum number of principal components was 3,determined by the minimum of prediction residual error sum of squares,and the prediction error of the regression equation ranges from -2.25% to 4.32%. The model predicted that the economic loss of vegetables from acid rain is negatively corrrelated to pH and the concentrations of NH4+,SO42-,NO3-,and Cl-in the rain,and positively correlated to the concentrations of Ca2+,Na+,K+ and Mg2+. The precision of the model may be improved if the non-linearity of original data is addressed.

基于PCA-HCA联合PLS回归模型的蚯蚓粪肥品质等级划分

蚯蚓粪肥理化特性涉及指标多,如何从众多易检测的指标中筛选出能够反映蚯蚓粪肥特点的关键指标,进而用于构建评价模型,高效、快速地评价蚯蚓粪肥的品质等级,是蚯蚓粪肥应用前亟需解决的重要问题与难点。研究针对不同原料类型、不同蚯蚓堆肥时间获得的蚯蚓粪肥,采用统计学与化学计量学对蚯蚓粪肥23个主要指标开展描述统计与相关分析,筛选出了13个蚯蚓粪肥特异性指标。以13个关键指标为基础,首先,结合主成分分析(PCA)与分层聚类分析(HCA)对不同蚯蚓粪肥样品开展品质初级划分;其次,采用偏最小二乘回归(PLS)-判别分析(DA)对分级结果进行效果判定;最后,整体构建基于PLS模型的蚯蚓粪肥等级评价方法并开展验证分析。结果表明:PCA与HCA分析法可将蚯蚓粪肥划分为3个品质等级,通过PLS-DA判别该划分结果合理有效,形成了基于PLS蚯蚓粪肥等级评价模型:蚯蚓粪肥品质等级(Y)=3.0796+0.0026×TOC-0.1381×HS-0.1446×HA-0.1378×TN-0.1355×TP-0.1494×AK-0.1324×AN-0.1402×AP+0.0004×EOC+0.03985×ROC+0.07685×C/N-0.0049×Kos-0.1481×HI(TOC、HS、HA、TN、TP、AK、AN、AP、EOC、ROC、C/N、Kos、HI分别代表总有机碳、腐殖质碳、胡敏酸、总氮、总磷、速效钾、碱解氮、有效磷、易氧化有机碳、难氧化有机碳、碳氮比、氧化稳定系数、腐殖化指数),分级标准为:若Y在0.45~1.56之间,品质等级为一等品;Y在1.63~2.20之间,为二等品;Y在2.28~3.72之间,为三等品。变量权重值表明影响蚯蚓粪肥品质前5的关键指标顺序为HI>TN>HS>HA>AN。研究成功建立了一套“PCA+HCA+PLS”的蚯蚓粪肥品质评价方法,对蚯蚓粪肥分级应用与规范蚯蚓产业市场具有重要意义。

基于递阶偏最小二乘回归的飞机采购价格预测

分析了飞机采购价格预测建模样本数据少、价格驱动因子众多的特点,考虑到递阶偏最小二乘回归(Hierarchical Partial Least-Squares Regression,Hi-PLS)方法在变量规模巨大情形下进行回归建模的优势,应用Hi-PLS对飞机采购价格进行预测。以战斗机采购价格预测为例进行了研究,首先对战斗机采购价格驱动因子进行分组,然后应用Hi-PLS对分组后的价格驱动因子进行回归,建立采购价格预测模型。实例表明,在飞机采购价格预测方面,采用递阶偏最小二乘回归预测更能体现价格与飞机性能参数之间的关系。

基于递阶偏最小二乘回归的数据分析

针对最小二乘法难以克服因子多重共线性对回归模型精度影响的不足和大坝观测数据分析中因变量较多的特征,引进递阶偏最小二乘法,对大坝安全监测变量及其影响因子进行递阶偏最小二乘回归分析,将建模预测分析方法通过递阶分层处理,可同时实现回归建模和数据结构简化,所建立的大坝安全监控模型精度可通过交叉有效性检验来控制.工程应用实例和模型对比分析研究表明,递阶偏最小二乘回归模型能有效克服由于各类因子变量间的多重共线性和因子变量数目较多而对模型拟合精度及其预测能力的影响,相对于传统回归模型有更好的解释能力,因而具有一定的实用价值.

基于小生境遗传算法的Hi-PLS回归模型优化

在以递阶偏最小二乘法建立回归模型的基础上,利用小生境遗传算法的快速、全局搜索优化功能对模型中的回归系数进行重新评估,建立了基于小生境遗传算法的递阶偏最小二乘大坝安全监控模型。工程实例与模型对比分析结果表明,经小生境遗传算法优化的递阶偏最小二乘回归模型不仅比原模型提高了拟合精度,且增强了预测能力。

一类非平坦函数的多核最小二乘支持向量机的鲁棒回归算法

给出了标准最小二乘支持向量机的数学回归模型,并提出了多核最小二乘支持向量机算法,用于提高非平坦函数的回归精度.运用谱系聚类方法解决多核最小二乘支持向量机的解缺乏稀疏性的问题.利用偏最小二乘回归方法对多核最小二乘支持向量机进行了鲁棒回归.通过仿真实例证实了所提方法的有效性.

递阶偏最小二乘回归在大坝安全监测中的应用

偏最小二乘回归能有效地消除因子间的多重相关性,但从其算法特点和实际应用来看,也存在不足。例如,在算法方面,偏最小二乘提取的主成分不一定能同时保证方差和相关程度最大;在应用方面,含有较多自变量的偏最小二乘回归模型的可解释性不高。递阶偏最小二乘回归是偏最小二乘回归后续研究的成果之一,在一定程度上克服了上述不足。算例表明,递阶偏最小二乘回归模型较其他回归模型的可解释性强,较为合理。

递阶偏最小二乘回归模型及其应用

偏最小二乘回归模型中包含所有原始选择的变量,当自变量较多时,因得到的模型结果十分庞杂而难以分析和解释。本文采用递阶偏最小二乘(Hierarchical PLS,Hi-PLS)回归方法,通过分层建立模型的方法有效解决了这一问题。工程实践表明,本模型精度较高,特别适用于大规模变量集合的回归分析。

基于新偏最小二乘回归法的系列水文资料分析

在实际问题中,经常会碰到海量数据或者样本点较少,自变量较多的数据。对此可以利用递阶偏最小二乘回归来建立线性模型。但是一个直接的问题是如何对自变量进行分组。由此提出了基于聚类分析的递阶偏最小二乘回归方法,在对解释变量分组时引入聚类分析。通过对长江宜昌段水沙观测数据作实证分析后发现,基于聚类分析的递阶偏最小二乘回归方法是有效可行的,而且用该方法建立的回归模型比一般的偏最小二乘回归模型拟合能力更强。

Multivariate analysis between meteorological factor and fruit quality of Fuji apple at different locations in China

China has the largest apple planting area and total yield in the world, and the Fuji apple is the major cultivar, accounting for more than 70％ of apple planting acreage in China. Apple qualities are affected by meteorological conditions, soil types, nutrient content of soil, and management practices. Meteorological factors, such as light, temperature and moisture are key environmental conditions affecting apple quality that are difficult to regulate and control. This study was performed to determine the effect of meteorological factors on the qualities of Fuji apple and to provide evidence for a reasonable regional layout and planting of Fuji apple in China. Fruit samples of Fuji apple and meteorological data were investigated from 153 commercial Fuji apple orchards located in 51 counties of 11 regions in China from 2010 to 2011. Partial least-squares regression and linear programming were used to analyze the effect model and impact weight of meteorological factors on fruit quality, to determine the major meteorological factors influencing fruit quality attributes, and to establish a regression equation to optimize meteorological factors for high-quality Fuji apples. Results showed relationships between fruit quality attributes and meteorological factors among the various apple producing counties in China. The mean, minimum, and maximum temperatures from April to October had the highest positive effects on fruit qualities in model effect loadings and weights, followed by the mean annual temperature and the sunshine percentage, the temperature difference between day and night, and the total precipitation for the same period. In contrast, annual total precipitation and relative humidity from April to October had negative effects on fruit quality. The meteorological factors exhibited distinct effects on the different fruit quality attributes. Soluble solid content was affected from the high to the low row preface by annual total precipitation, the minimum temperature from April to October, the mean temperature from April to October, the temperature difference between day and night, and the mean annual temperature. The regression equation showed that the optimum meteorological factors on fruit quality were the mean annual temperature of 5.5-18°C and the annual total precipitation of 602-1121 mm for the whole year, and the mean temperature of 13.3-19.6°C, the minimum temperature of 7.8-18.5°C, the maximum temperature of 19.5°C, the temperature difference of 13.7°C between day and night, the total precipitation of 227 mm, the relative humidity of 57.5-84.0％, and the sunshine percentage of 36.5-70.0％ during the growing period （from April to October）.

Optimizing process of preparing artificial-similar material for rocky slope with uniform formula design

The determination of material formula needs try-and-error experiment,and consumes large amount of time and fund.In order to solve the problem,a comprehensive method is established,via the experiment of artificial-similar material formula of a mine slope.We controlled the samples by the compactness,and arranged the formula of the test group with the method of the uniform formula experiment.The physical and mechanical parameters of these samples were analyzed using the method of the partial least-squares regression(PLS).And a mathematical model of the indexes of physical and mechanics parameters relating to the factors of formulation constituents was established eventually.We used the model to analyze the effect of each formulation constituent on physical and mechanics parameters of samples.The experiment results and analysis illustrates that1)in the formulation of similar material,the effect of raw materials on the internal friction angleφand cohesion C is opposite;2)The method can highly facilitate the process of the of preparing artificial-similar materials,more economic and effective.

Studies on a Novel Characteristic Atom-pair Holographic Code Applied to Quantitative Structure-chromatographic Retention Relationship of Organic Compounds

6 Atomic fragment types of organic compound have been defined, and the multilevel atom-pair frequency matrix has been constructed according to the occurrence number in pairs of atomic fragments with different bond lengths in the molecule. On the basis of them, a novel molecular coding technique： characteristic atom-pair holographic code （CAHC）, is obtained. To some extent, this method exhibits a large number of benefits at the same time. For example, it can calculate 2D molecular topological descriptor easily, operate without difficulty and possess definite physicochemical meaning of 3D molecular structural characterization methods, and may fetch the complicated information of molecule, etc. Therefore, it is appropriate for the study on quantitative structure-property/activity relationship （QSPR/QSAR） of medicines and biological molecules. We attempt in this paper to utilize the method of CAHC to the quantitative prediction of reversed-phase liquid chromatogram （RPLC） retention data of 33 purine derivatives and 24 steroids. The fitting multiple correlation coefficient R2, cross-validated multiple correlation coefficient Q2 and predicted ability Q^2 pred over test set＇s samples of obtained partial least-square （PLS） regression model are respectively 0.990, 0.893 and 0.977, 0.897, 0.941.

Simultaneous Determination of Verapamil Hydrochloride and Gliclazide in Synthetic Binary Mixture and Combined Tablet Preparation by Chemometric-Assisted Spectroscopy

In this study, the simultaneous determination of verapamil hydrochloride and gliclazide in pharmaceuticals by chemometric approaches using UV spectrophotometry has been reported. Verapamil hydrochloride (VER) (Benzeneacetonitrile, α-[3-[[2-(3,4-dimethoxyphenyl) ethyl] methylamino]propyl]-3, 4-dimethoxy-α-(1-methylethyl) hydrochloride) is an L-type calcium channel blocker of the phenylalkylamine class. It has been used in the treatment of hypertension, angina pectoris, and cardiac arrhythmia. Gliclazide (GLZ) (1-(Hexahydrocyclopenta[c]pyrrol-2(1H)-yl)-3-[(4-methylphenyl) sulphonyl]urea) is an oral hypoglycaemic (anti-diabetic) drug and is classified as a second generation sulfonylurea. Spectra of VER and GLZ were recorded at several concentrations within their linear ranges between wavelengths of 200 nm to 400 nm in 0.1N HCl. Partial least squares regression (PLS) and principle components regression (PCR) were used for chemometric analysis of data and the parameters of the chemometric procedures were optimized. The recoveries were satisfactory and statistically comparable. The method was successfully applied to pharmaceutical formulation, tablet, with no interference from excipients as indicated by the recovery study results. The proposed methods are simple, rapid and can be easily used in the quality control of drugs as alternative analysis tools.

递阶偏最小二乘回归在飞机研制费用预测中的应用

分析了飞机研制费用样本数据少、费用驱动因子众多的特点,考虑到递阶偏最小二乘回归(Hi-PLS)方法在变量规模巨大情形下进行回归建模的优势,应用递阶偏最小二乘回归方法对飞机研制费用进行预测。以战斗机机体研制费用预测为例进行研究,首先对战斗机机体研制费用驱动因子进行分组,然后应用递阶偏最小二乘回归方法对分组后的费用驱动因子进行回归建立机体研制费用预测模型。实例表明,在飞机研制费用预测方面,采用递阶偏最小二乘回归方法预测更能体现研制费用与飞机性能参数之间的关系。

女儿茶HPLC指纹图谱及清除自由基活性谱效关系

目的：研究女儿茶Rhamnus heterophylla HPLC指纹图谱与其清除自由基活性的关系。方法：采用高效液相色谱法建立8批不同产地、采收期女儿茶指纹图谱,在传统相似度评价基础上,采用Origin Pro 2017软件进行层次聚类分析（hierarchical cluster analysis,HCA）和主成分分析（principal component analysis,PCA）对其指纹图谱中的共有峰进行评价;1,1-二苯基-2-三硝基苯肼（DPPH）法研究其清除自由基活性;偏最小二乘回归分析（partial least squares regression,PLSR）和灰色关联度分析（grey relational analysis,GRA）研究谱效关系。结果：建立了8批女儿茶HPLC指纹图谱,确定了11个共有峰,相似度在0.948-0.976。采用对照品比对方法指认了其中6个峰：x1为原儿茶酸,x2为木犀草苷,x5为槲皮素,x8为山柰酚,x10为大黄素8-O-α-L-鼠李糖苷,x11为大黄素。样本可聚为4类。8批女儿茶均有不同程度清除自由基活性。灰色关联度分析结果显示共有峰与清除自由基活性的关联度大小为x11〉x10〉x8〉x1〉x9〉x4〉x3〉x2〉x6〉x5〉x7。偏最小二乘回归分析建立的回归方程为Y=99.769 17-0.357 49x1-0.001 36x2-0.002 65x3＋0.059 63x4＋0.011 81x6＋0.010 63x7-0.006 99x8-0.009 14x9＋0.054 27x10＋0.022 75x11。结论：女儿茶清除自由基活性是多组分联合效应的结果。

基于模式识别和遗传神经网络算法的醋香附近红外光谱等级评价和含量预测模型研究

目的基于近红外光谱(near infrared spectrum,NIRS)技术建立一种能快速准确识别醋香附饮片等级并预测其挥发油中α-香附酮、香附烯酮含量的质量评价模型,为其他中药材或中药饮片的质量评价提供参考。方法采集醋香附的NIRS信息,并建立39批醋香附挥发油气相色谱-质谱联用(GC-MS)指纹图谱,对挥发油中的α-香附酮、香附烯酮进行定量,采用相似度分析、多元统计分析、主成分分析(principal component analysis,PCA)、聚类分析、偏最小二乘-判别分析(partial least squares-discriminant analysis,PLS-DA)、Logistic回归分析等方法处理数据,划分等级;利用遗传神经网络算法(GA-BPNN)将等级划分结果、α-香附酮含量、香附烯酮含量分别与NIRS信息进行拟合,建立等级预测模型和含量预测模型。结果根据主成分聚类分析法可以将醋香附划分为3个等级,其中一等品6批,二等品8批,三等品25批,PLS-DA分析结果与主成分聚类分析结果一致。采用多元Logistic回归分析建立了饮片等级分类经验公式P一等=exp(G1)/[1+exp(G1)]、P二等=exp(G2)/[1+exp(G2)]、P三等=1-P二等,等级预测结果和主成分聚类分析结果一致。利用GA-BPNN建立的醋香附饮片等级预测模型预测准确率达89.74%,模型准确性较好;α-香附酮、香附烯酮回归模型预测集决定系数分别为0.9923、0.9697,能很好地预测醋香附挥发油中α-香附酮、香附烯酮含量。结论采用GA-BPNN所建立的基于近红外技术的醋香附饮片等级评价模型能快速准确地预测醋香附饮片等级,为醋香附及其他中药材或中药饮片质量标准的制定和等级评价模型的研究提供了参考。

Discrimination of Chuanminshen violaceum Sheh et Shen from different regions based on fatty acid profiles of roots and leaves

Objectives:The purpose of this paper was to construct a reliable methodology to discriminate the geographical origins of Chuanminshen violaceum Sheh et Shan planted in different regions in Sichuan,China.Materials and methods:Fatty acid profiles of roots and leaves of C.violaceum planted in various regions of Sichuan Province in China,namely Guangyuan(GY),Langzhong(LZ),Jintang(JT),Bazhong(BZ),and Shuangling(SL),were determined using GC-MS followed by multivariate statistical analyses,including orthogonal partial least-squares discriminant analysis and hierarchical clustering analysis.Results:Leaves of C.violaceum showed the highest contents of hexadecatrienoic acid(3.21 g/kg),linoleic acid(6.62 g/kg),andα-linolenic acid(7.24 g/kg),which were all higher than those contained in roots.Chuanminshen violaceum samples collected from LZ,JT,and GY could be clearly distinguished based on fatty acid profiles of leaves and those collected from LZ,GY,and BZ could be clearly distinguished based on fatty acid profiles of roots.Conclusions:Chemometric method is used as a potential approach for analyses of fatty acid profiles of roots and leaves to control the quality of C.violaceum and their powered products.