A novel methodology for investigating foamy oil stability by an oil-based analogue model

A novel methodology was developed based on oil-based foam to systematically investigate foamy oil stability. The first step was to generate an oil-based analogue model that has similar characteristics of foamy oil. Then, the effects of the key factors were investigated with the created oil-based foam by a new experimental procedure. Moreover, a statistical approach was utilized to ensure reliability of the experimental data. Finally, we proposed a new method to describe foamy oil rupture and simulate its defoaming kinetics. The results showed that the oil-based foam can be precursors to studies on actual foamy oil and other petroleum foams.     

A tutorial on support vector machine-based methods for classification problems in chemometrics

This tutorial provides a concise overview of support vector machines and different closely related techniques for pattern classification. The tutorial starts with the formulation of support vector machines for classification. The method of least squares support vector machines is explained. Approaches to retrieve a probabilistic interpretation are covered and it is explained how the binary classification techniques can be extended to multi-class methods. Kernel logistic regression, which is closely related to iteratively weighted least squares support vector machines, is discussed. Different practical aspects of these methods are addressed: the issue of feature selection, parameter tuning, unbalanced data sets, model evaluation and statistical comparison. The different concepts are illustrated on three real-life applications in the field of metabolomics, genetics and proteomics.     

Sequential estimation in regression models using analogues of trimmed means

A sequential procedure is proposed for constructing a fixed-size confidence region for the parameters of a linear regression model. The procedure is based on certain regression analogues of trimmed means, as formulated by Welsh (1987,Ann. Statist.,15, 20–36), rather than least squares estimates. For error distributions with continuous, symmetric density and some moment higher than fourth finite, if the design points of the model are bounded, then the procedure is both asymptotically consistent and asymptotically efficient as the size of the region approaches zero.Research supported in part by the National Science Foundation under Grants DMS 85-03321 and 88-02556 and by the Air Force under Grant AFOSR-87-0041.     

Melamine detection by mid- and near-infrared (MIR/NIR) spectroscopy: a quick and sensitive method for dairy products analysis including liquid milk, infant formula, and milk powder

Melamine (2,4,6-triamino-1,3,5-triazine) is a nitrogen-rich chemical implicated in the pet and human food recalls and in the global food safety scares involving milk products. Due to the serious health concerns associated with melamine consumption and the extensive scope of affected products, rapid and sensitive methods to detect melamine's presence are essential. We propose the use of spectroscopy data-produced by near-infrared (near-IR/NIR) and mid-infrared (mid-IR/MIR) spectroscopies, in particular—for melamine detection in complex dairy matrixes. None of the up-to-date reported IR-based methods for melamine detection has unambiguously shown its wide applicability to different dairy products as well as limit of detection (LOD) below 1 ppm on independent sample set. It was found that infrared spectroscopy is an effective tool to detect melamine in dairy products, such as infant formula, milk powder, or liquid milk. ALOD below 1 ppm (0.76 ± 0.11 ppm) can be reached if a correct spectrum preprocessing (pretreatment) technique and a correct multivariate (MDA) algorithm—partial least squares regression (PLS), polynomial PLS (Poly-PLS), artificial neural network (ANN), support vector regression (SVR), or least squares support vector machine (LS-SVM)—are used for spectrum analysis. The relationship between MIR/NIR spectrum of milk products and melamine content is nonlinear. Thus, nonlinear regression methods are needed to correctly predict the triazine-derivative content of milk products. It can be concluded that mid- and near-infrared spectroscopy can be regarded as a quick, sensitive, robust, and low-cost method for liquid milk, infant formula, and milk powder analysis.     

求解大规模矛盾方程组的最小二乘支持向量机算法

房价预测、共享单车出租数量预测、空气污染情况预测等常涉及矛盾方程组求解,对其数值求解方法研究具有重要的理论意义与应用价值。当矛盾方程组规模过大时,用传统的最小二乘法求解,不仅计算量大,而且由于误差积累使最终结果的准确性不高。鉴于此,采用机器学习中的最小二乘支持向量机（least squares support vector machine,LS-SVM）算法求解大规模矛盾方程组,并分别针对线性、非线性、单变量、多变量矛盾方程组进行了数值求解。数值结果表明,数据类型和数据量的变化对结果的影响不大,因此只要选取适当的参数就可建立合适的模型,得到高精度的预测结果。     

应用高光谱成像技术鉴别绿茶品牌研究

应用高光谱成像技术,基于光谱主成分信息和图像信息的融合实现名优绿茶不同品牌的鉴别。首先采集6个品牌名优绿茶在380～1 023 nm波长范围的512幅光谱图像,然后提取并分析绿茶样本的可见近红外光谱响应特性,结合主成分分析法找到了最能体现这6类样本差异的2个特征波段(545和611 nm),并从这2个特征波段图像中分别提取12个灰度共生矩阵纹理特征参量包括中值、协方差、同质性、能量、对比度、相关、熵、逆差距、反差、差异性、二阶距和自相关,最后融合这12个纹理特征和三个主成分特征变量得到名优绿茶品牌识别的特征信息,利用LS-SVM建立区分模型,预测集识别率达到了100%,同时采用ROC曲线的评估方法来评估分类模型。结果表明综合应用灰度共生矩阵变量和光谱主成分变量作为LS-SVM模型输入可实现对绿茶品牌的鉴别。     

基于LS-SVM和THz光谱技术的面粉中苯甲酸检测研究

面粉（小麦粉）是中国北方大部分地区的主食,苯甲酸是重要的酸型食品防腐剂,为了便于食品长期保存,往往会添加苯甲酸以便延长食品保存时间。但食用添加苯甲酸过量的小麦粉会对身体健康产生严重危害。太赫兹技术是一种新兴的检测技术,由于处于特殊的0.1～10 THz的太赫兹频段,在食品安全检测方面体现出了很强的应用潜力。主要致力于探索太赫兹光谱技术检测苯甲酸的合理性、可行性,利用太赫兹时域光谱技术对面粉中的食品添加剂苯甲酸进行实验研究。实验获取了面粉和苯甲酸的太赫兹时域光谱和频域光谱,其吸收系数显示苯甲酸的特征吸收峰在1.94 THz波段,面粉的太赫兹吸收系数几乎以一定的斜率增加,说明可以用THz-TDS（Terahertz time domain spectrum）技术对面粉中的苯甲酸进行特征识别。为建立面粉中添加剂苯甲酸的定量检测模型,实验获取了面粉中掺杂不同百分比（质量分数）苯甲酸的太赫兹时域光谱,计算得到吸收系数谱。实验发现吸收峰幅度的变化是与苯甲酸的含量成正比的,苯甲酸含量增加吸收峰幅度变大。首先探索了不同光谱预处理方法对太赫兹光谱的影响,采用如平滑校正、多元散射校正、基线校正和归一化等方法对原始光谱进行校正处理。校正之后,建立相应的PLS (partial least squares)模型以选择最优预处理方法。然后分别建立苯甲酸浓度和太赫兹吸收系数的MLR (multiple linear regression)、PLS和LS-SVM（partial least squares support vector machines）回归模型,并对比分析不同模型的优劣。将光谱数据归一化后建立的PLS模型更具有优势,预测相关系数R_p为0.979,预测均方根误差RMSEP为1.30%。LS-SVM与PLS和MLR模型相比,LS-SVM模型可以获得更好的建模结果,LS-SVM的预测相关系数R_p为0.987,预测均方根误差RMSEP为1.10%。利用MLR方法仅使用1.946和1.869 THz两个波段点进行建模,建模效果预测相关系数R_p为0.955,预测均方根误差RMSEP为1.90%。通过该研究为面粉中苯甲酸添加剂的无损检测提供了新的解决方案,也为其他类型的添加剂的检测提供了方法指导,对促进面粉行业的健康发展具有重要的意义。     

中红外光谱对甲醇汽油甲醇含量检测研究

甲醇汽油是一种清洁能源,甲醇汽油中甲醇的含量决定了汽油的性能。通过中红外光谱对甲醇汽油中甲醇含量进行定量检测和分析。首先,对采集的甲醇汽油原始中红外光谱进行平滑处理（smoothing）、多元散射校正（MSC）、基线校正（baseline）、归一化（normalization）等预处理,再建立PLS模型,对比选择最佳预处理方法,结果表明：在多元散射校正（MSC）处理后建立的PLS模型效果最好,模型的预测集相关系数r为0.918,预测均方根误差RMSEP为2.107。为进一步简化模型,提高预测精度,采用无信息变量消除(uninformative variable elimination, UVE)方法对波长进行筛选,将UVE波段筛选之后的作为模型的输入变量,采用偏最小二乘法（partial least squares, PLS）、主成分回归(principal components regression, PCR)和最小二乘支持向量机（least square support vector machine, LSSVM）三种方法分别建立甲醇汽油中甲醇含量的定量预测模型,并比较不同模型的预测效果和结果。结果表明,使用无信息变量消除可以较好提高数据的运算速度,其中,UVE-PLS模型建模效果最好,r和RMSEP分别为0.923和2.075。该实验表明中红外光谱检测甲醇汽油中甲醇含量是可行的并可以得到较好的效果;UVE是一种对甲醇汽油的中红外光谱非常有效的波段筛选方法,该模型的建立对石油化工领域具有较为重要的意义。     

基于紫外光谱的水体硝酸盐浓度混合预测模型研究

水体中的硝酸盐浓度过高不仅会造成水环境污染而且会对人类身体健康造成很大威胁,传统的检测硝酸盐的方法检测时间长且操作复杂。针对水体中硝酸盐氮难以快速在线检测的问题,基于紫外吸收光谱,提出了一种混合预测模型结合光谱积分快速定量检测水体中硝酸盐浓度的方法。混合预测模型为低浓度样本建立的双波长法预测模型与高浓度样本建立的偏最小二乘支持向量机(LS-SVM)预测模型数据融合之后的模型。按照合适的浓度梯度配备了19组硝酸盐氮标准溶液,通过实验测得不同浓度硝酸盐氮样本的光谱数据。首先基于双波长法对所有样本进行回归分析,按照A=A₂₂₀-2A₂₇₅计算不同实验样本的吸光度A,其中A₂₂₀和A₂₇₅是220和275 nm处样本的吸光度,将吸光度A与样本浓度值进行线性回归,拟合出样本浓度的预测值。结果显示当样本浓度较小时,相关性很好,r为0.997 4,随着实验样本浓度的上升,曲线发生严重的非线性漂移,因此双波长法只适合低浓度样本预测模型的建立。对于高浓度样本,光谱重叠严重,适合建立非线性的预测模型,支持向量机(SVM)与LS-SVM都适合小样本的非线性数据建模,LS-SVM预测精度稍高,运算速度稍快。通过对所有的实验样本进行全波长光谱积分,比较相邻样本光谱积分的变化率可以筛选出样本的临界浓度值,4 mg·L-1的硝酸盐样本积分值前后变化率最大,因此选择4 mg·L-1作为临界浓度值较为合适。浓度高于4 mg·L-1的实验样本建立LS-SVM预测模型,通过交叉验证的方法选择出合适的参数,正则化参数γ=50,核函数选择高斯核,核函数宽度σ２=0.36,训练样本之后进行回归;其余样本建立双波长法预测模型,最后进行两种模型的数据融合,形成从低浓度到高浓度的水体中硝酸盐浓度的检测。为了验证混合预测模型的预测精度,另外建立了SVM,LS-SVM,偏最小二乘(PLS)等模型,并求出r,预测值与真实浓度值平均绝对误差(MAE)和均方根误差(RMSE)对模型进行评价。验证结果表明,相比于SVM,LS-SVM和PLS等模型,提出的混合模型回归的相关系数为0.999 86,分别提高了1.8%,1.6%和0.45%,预测值与真实浓度的平均绝对误差为2.55%,分别降低了6.27%,4.49%和1.01%,均方根误差为0.303,为四种预测模型中最小,SVM与LS-SVM的相对误差相对较高,PLS模型相对误差上下波动比较大,混合预测模型相对误差最为稳定,并保持在较低水平,由此可见混合预测模型的预测效果明显优于其他几种模型。并与文献[5-7]中的测量方法进行对比,该混合预测方法可以简单快速的测量水体中硝酸盐氮的浓度,且不需要试剂,无二次污染,与文献[9]中的预测模型相比,预测精度明显提高。因此提出的混合模型可正确快速地预测水体中硝酸盐氮的浓度,可为在线监测水体中硝酸盐浓度提供有效的技术参考。