PCA-LVQ法及其在RS-FTIR大气环境监测数据处理中的应用

将主成分分析(PCA)用于遥感傅里叶变换红外光谱(Remote Sensing Fourier Transform Infrared:RS-FTIR)的特征提取,结合学习矢量量化(LVQ)神经网络,实现了PCA-LVQ对大气中的8组分混合体系进行快速定性分析的建模方法。并与单纯的LVQ神经网络、反向传播人工神经网络(BP-ANN)得到的结果进行了比较。PCA-LVQ显示出较好的处理数据的能力,它不仅提高了运算速度,而且提高了模型的预测准确度,分类精度达到91.7%。PCA-LVQ的这一预测精度及运算速度,足以满足遥感傅里叶变换红外光谱对大气中有毒气体的实时、在线监测的需要。     

Near-infrared (NIR) spectroscopy for motor oil classification: From discriminant analysis to support vector machines

Near-infrared (NIR) spectroscopy is a non-destructive measurement technique for many chemical compounds that has proved its efficiency for laboratory and industrial applications (including petroleum industry). Motor oil classification is an important task for quality control and identification of oil adulteration. Type of motor oil base stock is a key factor in product price formation. In this paper we have tried to evaluate the efficiency of different methods for motor oils classification by base stock (synthetic, semi-synthetic and mineral) and kinematic viscosity at low and high temperature. We have compared the abilities of seven (7) different classification methods: regularized discriminant analysis (RDA), soft independent modelling of class analogy (SIMCA), partial least squares classification (PLS), K-nearest neighbour (KNN), artificial neural network - multilayer perceptron (ANN-MLP), support vector machine (SVM), and probabilistic neural network (PNN) - for classification of motor oils. Three (3) sets of near-infrared spectra (1125, 1010, and 1050 items) were used for classification of motor oils into three or four classes. In all cases NIR spectroscopy was found to be effective for motor oil classification when combined with an effective multivariate data analysis (MDA) technique. SVM and PNN chemometric techniques were found to be the most effective ones for classification of motor oil based on its NIR spectrum.     

基于小波神经网络的新型算法用于化学信号处理

基于紧支集正交小波神经网络的构造思想,用具有紧支集的Ｂ－样条函数的伸缩和平稳替代小波函数,提出了一种新型算法,并将其应用于化学信号的处理,实现了信号的压缩和滤噪,适应小波神经网络相比,学习速度得到了大幅度的提高。     

Evaluation of chemometric techniques and artificial neural networks for cancer screening using Cu, Fe, Se and Zn concentrations in blood serum

It is known that variations in the concentrations of certain trace elements in bodily fluids may be an indication of an alteration of the organism's normal functioning. Multivariate analysis of such data, and its comparison against proper reference values, can thus be employed as possible screening tests. Such analysis has usually been conducted by means of chemometric techniques and, to a lower extent, backpropagation artificial neural networks. Despite the excellent classification capacities of the latter, its development can be time-consuming and computer-intensive. Probabilistic artificial neural networks represent another attractive, yet scarcely evaluated classification technique which could be employed for the same purpose. The present work was aimed at comparing the performance of two chemometric techniques (principal component analysis and logistic regression) and two artificial neural networks (a backpropagation and a probabilistic neural network) as screening tools for cancer. The concentrations of copper, iron, selenium and zinc, which are known to play an important role in body chemistry, were used as indicators of physical status. Such elements were determined by total reflection X-ray fluorescence spectrometry in a sample of blood serum taken from individuals who had been given a positive (N = 27) or a negative (N = 32) diagnostic on various forms of cancer. The principal components analysis served two purposes: (i) initial screening of the data; and, (ii) reducing the dimension of the data space to be input to the networks. The logistic regression, as well as both artificial neural networks showed an outstanding performance in terms of distinguishing healthy (specificity: 90-100%) or unhealthy individuals (sensitivity: 100%). The probabilistic neural network offered additional advantages when compared to the more popular backpropagation counterpart. Effectively, the former is easier and faster to develop, for a smaller number of variables has to be optimized, and there are no risk of overtraining.     

大鼠胰腺及癌组织红外光谱连续小波特征提取及径向基人工神经网络识别

采用水平衰减全反射(HATR)傅里叶变换红外光谱法(FTIR)测定了SD大鼠胰腺正常组织与非正常组织的谱图,提出了一种新的基于FTIR的连续小波特征提取与径向基人工神经网络分类方法以提高FTIR对早期SD大鼠胰腺癌的诊断准确率。利用连续小波多分辨率分析法提取FTIR特征量,对于提取的特征量采用径向基函数神经网络进行模式分类。对SD大鼠的胰腺正常组织、早期癌组织及进展期癌组织的FTIR,利用连续小波多分辨率分析法提取9个特征量,进行RBF神经网络分类判断。当目标误差为0.01,径向基函数的分布常数为5时,网络达到最优化,总的正确识别率为96.67%。并对影响分类结果的网络参数、目标误差和分布常数对分类样品的影响做了讨论。实验结果表明:此方法对早期胰腺癌具有较高的诊断率。     

Learning vector quantization for multiclass classification: application to characterization of plastics

Learning vector quantization (LVQ) is described, with both the LVQ1 and LVQ3 algorithms detailed. This approach involves finding boundaries between classes based on codebook vectors that are created for each class using an iterative neural network. LVQ has an advantage over traditional boundary methods such as support vector machines in the ability to model many classes simultaneously. The performance of the algorithm is tested on a data set of the thermal properties of 293 commercial polymers, grouped into nine classes: each class in turn consists of several grades. The method is compared to the Mahalanobis distance method, which can also be applied to a multiclass problem. Validation of the classification ability is via iterative splits of the data into test and training sets. For the data in this paper, LVQ is shown to perform better than the Mahalanobis distance as the latter method performs best when data are distributed in an ellipsoidal manner, while LVQ makes no such assumption and is primarily used to find boundaries. Confusion matrices are obtained of the misclassification of polymer grades and can be interpreted in terms of the chemical similarity of samples.     

Comparison of supervised pattern recognition methods with McNemar’s statistical test: Application to qualitative analysis of sugar beet by near-infrared spectroscopy

The application of supervised pattern recognition methodology is becoming important within chemistry. The aim of the study is to compare classification method accuracies by the use of a McNemar’s statistical test. Three qualitative parameters of sugar beet are studied: disease resistance (DR), geographical origins and crop periods. Samples are analyzed by near-infrared spectroscopy (NIRS) and by wet chemical analysis (WCA). Firstly, the performances of eight well-known classification methods on NIRS data are compared: Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN) method, Soft Independent Modeling of Class Analogy (SIMCA), Discriminant Partial Least Squares (DPLS), Procrustes Discriminant Analysis (PDA), Classification And Regression Tree (CART), Probabilistic Neural Network (PNN) and Learning Vector Quantization (LVQ) neural network are computed. Among the three data sets, SIMCA, DPLS and PDA have the highest classification accuracies. LDA and KNN are not significantly different. The non-linear neural methods give the less accurate results. The three most accurate methods are linear, non-parametric and based on modeling methods. Secondly, we want to emphasize the power of near-infrared reflectance data for sample discrimination. McNemar’s tests compare classification developed with WCA or with NIRS data. For two of the three data sets, the classification results are significantly improved by the use of NIRS data.     

Detection of Cholangiocarcinoma with Fourier Transform Infrared Spectroscopy and Radial Basis Function Neural Network Classification

The aim of this study was to explore the possibility of applying Fourier transform infrared(FTIR) spectroscopy as a medical diagnostic tool based on a neural network classifier for detecting and classifying cholangiocarcinoma. A total of 51 cases of bile duct tissues were obtained and later characterized by FTIR spectroscopy prior to pathological diagnosis. The criteria for classification included 30 parameters for each FTIR spectra, including peak position(P), intensity(I) and full width at half-maximum(FWHM), were measured, calculated and subsequently compared against the normal and cancer groups. The FTIR spectra were classified by the radial basis function(RBF) network model. For establishing the RBF, 23 cases were used to train the RBF classifier, and 28 cases were applied to validate the model. Using the RFB model, nine parameters were observed to be pronouncedly different between cancerous and normal tissue, including I₁₆₄₀, I₁₅₅₀, I₁₄₆₀, I₁₄₀₀, I₁₂₅₀, I₁₁₂₀, I₁₀₈₀, I₁₀₄₀ and P₁₀₄₀. In the RBF training classification, the accuracy, sensitivity, and specificity of diagnosis were 82.6%, 80.0%, and 84.6%, respectively. While validating the classification, the accuracy, sensitivity, and specificity of diagnosis were 78.6%, 75.0%, and 81.2%, respectively. The results suggest that FTIR spectroscopy combined with neural network classifier could be applied as a medical diagnostic tool in cholangiocarcinoma diagnosis.     

Combinative neural network and its applications

A new approach named combinative neural network (CN) using partial least squares (PLS) analysis to modify the hidden layer in the multi-layered feed forward (MLFF) neural networks (NN) was proposed in this paper. The significant contributions of PLS in the proposed CN are to reorganize the outputs of hidden nodes such that the correlation of variables could be circumvented, to make the network meet the non-linear relationship best between the input and output data of the NN, and to eliminate the risk of over-fitting problem at the same time. The performance of the proposed approach was demonstrated through two examples, a well defined nonlinear approximation problem, and a practical nonlinear pattern classification problem with unknown relationship between the input and output data. The results were compared with those by conventional MLFF NNs. Good performance and time-saving implementation make the proposed method an attractive approach for non-linear mapping and classification.     

Classification of Rat FTIR Colon Cancer Data Using Waveletsand BPNN

本文提出了一种新的基于水平衰减全反射-傅里叶变换红外光谱（HATR-FTIR）的小波特征提取与反向传播人工神经网络模式分类方法以提高FTIR对早期大鼠结肠癌的诊断准确率．对60只DMH诱导的SD大鼠,44只诱导鼠的第二代鼠,36只正常SD大鼠的结肠正常组织、异常增生、早癌及进展期癌组织所获得的的HATR-FTIR,利用连续小波多尺度分析法提取12个特征量,采用反向传播人工神经网络进行分类,识别准确率分别为100%、94%、97.5%及100%．实验结果表明此方法对早期结肠癌具有较高的诊断率．     

级联神经网络用于单扫描示波极谱信号的测定

将改进小波神经网络与BP神经网络相结合,提出一种新的混级联神经网络结构,用于单扫描示波极谱信号的同时测定。通过对网络结构的优化和网络参数的调整,加快了训练速度,提高了预测的准确度。用该法对邻、间硝基氯苯混合样进行了预测,结果满意。对级联神经网络法与单一BP神经网络法的预测结果进行了比较,表明级联神经网络优于单一BP神经网络。     

灰铸铁石墨形态的自动分类

在所提取的纹理特征的基础上,使用误差后向传播神经网络构建了一种优化的人工神经网络人顺。实现了灰铸铁石墨态的自动分类。用于描述石墨形态特征由分形维,粗细参数和二维自回归系数共同组成。该法成功地将人工神经网络引入了对灰铸铁石墨形态的分类,相对于传统人工目测法是一种很大的进步,而神经网络分类器的优化方法对其它神经网络模型的构也具有一定参考价值。     

Predicting cis/trans structure of alkene based on infrared spectra.

The application of chemometrics to analyze the information of the cis/trans structure of alkenes in infrared spectra (IR) is introduced. For data from the OMNIC IR spectral database, two feature selection methods, Fisher ratios and genetic algorithm-partial least squares (GA-PLS), and two classification methods, support vector machine (SVM) and probabilistic neural network (PNN), have been used to obtain optimization classifiers. At last, some spectra from other IR databases are used to evaluate the optimization classifiers. It has been demonstrated that both the SVM and PNN optimization classifiers could give preferable predictive results about the cis and trans structures of alkene.     

盐湖水化学类型的人工神经网络判别方法

研究了作为典型径向基函数网络之一的概率神经网络在盐湖水化学类型分类预测中的应用，验证了该方法的可靠性，得到了满意的分类预测结果。实验结果和网络结构分析表明，概率神经网络方法比熟知的反向传播算法(BP)网络要好。概率神经网络的研究应用为化学模式识别提供了一个新工具。     

Classification of Chinese Herbal Medicine by Laser-Induced Breakdown Spectroscopy with Principal Component Analysis and Artificial Neural Network

Chinese herbal medicine has attracted increasing attention because of the unique and significant efficacy in various diseases. In this paper, three types of Chinese herbal medicine, the roots of Angelica pubescens, Codonopsis pilosula, and Ligusticum wallichii with different places of origin or parts, are analyzed and identified using laser-induced breakdown spectroscopy (LIBS) combined with principal component analysis (PCA) and artificial neural network (ANN). The study of the roots of A. pubescens was performed. The score matrix is obtained by principal component analysis, and the backpropagation artificial neural network (BP-ANN) model is established to identify the origin of the medicine based on LIBS spectroscopy of the roots of A. pubescens with three places of origin. The results show that the average classification accuracy is 99.89%, which exhibits better prediction of classification than linear discriminant analysis or support vector machine learning methods. To verify the effectiveness of PCA combined with the BP-ANN model, this method is used to identify the origin of C. pilosula. Meanwhile, the root and stem of L. wallichii are analyzed by the same method to distinguish the medicinal materials accurately. The recognition rate of C. pilosula is 95.83%, and that of L. wallichii is 99.85%. The results present that LIBS combined with PCA and BP-ANN is a useful tool for identification of Chinese herbal medicine and is expected to achieve automatic real-time, fast, and powerful measurements.     

一类基于模糊神经元的复杂非线性化学模式识别方法

针对模式类别边界曲折而模糊的复杂化学模式分类问题,提出一种化学模式模糊分类方法,并给出其模糊神经元分类器设计和网络训练算法,使模糊神经元分类器具有学习功能.以一个应用实例检验了该方法的实效.     

用于药品质量快速检测的近红外光谱模糊神经元分类方法

针对非线性且分类界线模糊的药品质量类别快速测定难题, 将近红外光谱分析与模糊神经网络相结合, 经研究提出近红外光谱模糊神经网络分类方法, 用于计算辨析中药等化学组成复杂药品的近红外光谱模式类别, 从而快速评定这类药品的质量. 以参麦注射液为典型分析对象, 以鉴别其生产厂家这一模式分类问题为例, 考核本文方法, 结果表明, 其分类准确率达到94.2%, 明显优于经典的BP神经网络分类方法(84.6%), 可望用于中药产品质量类别的快速检测与评价.     

人工神经网络在溶解度预测方面的应用

溶解度作为一项重要的物化指标,一直是化学学科的研究重点。然而,通过实验测量获得数据耗时费力,因此,科研人员建立了多种理论方法来进行估算,其中,人工神经网络因其能够关联复杂的多变量情况而受到广泛关注。本文综述了人工神经网络在物质溶解度预测方面的应用,介绍了应用最广泛的3种神经网络(BP神经网络、小波神经网络、径向基神经网络)的模型结构、预测方法和预测优势,探讨了神经网络的不足以及改进方法。文章最后对神经网络在物质溶解度预测方面的发展前景进行了展望。与其他方法相比,人工神经网络技术在物质溶解度预测方面具有预测结果精确度高、操作简单等特点,具有广阔的应用前景,但输入变量选择、隐含层节点数确定、避免局部最优等问题还需逐步建立系统的理论指导。