基于电子鼻系统的番茄苗不同种类损伤的区分效果研究

通过对早疫病病害番茄苗、灰霉病病害番茄苗、机械损伤番茄苗和对照番茄苗的电子鼻响应信号的对比,可以看出不同处理的番茄苗样本电子鼻的响应信号是不同的,表明用电子鼻响应信号对番茄苗不同种类损伤进行预测是可行的.从PCA结果来看,早疫病病害的番茄苗和灰霉病病害的番茄苗能很好区分开,机械损伤的番茄苗和正常处理的番茄苗产生了重叠现...     

Enhanced iterative projection for subclass discriminant analysis under EM-alike framework

Mixture discriminant analysis (MDA) and subclass discriminant analysis (SDA) belong to the supervised classification approaches. They have advantage over the standard linear discriminant analysis (LDA) in large sample size problems, since both of them divide the samples in each class into subclasses which keep locality but LDA does not. However, since the current MDA and SDA algorithms perform subclass division in just one step in the original data space before solving the generalized eigenvalue problem, two problems are exposed: (1) they ignore the relation among classes since subclass division is performed in each isolated class; (2) they cannot guarantee good performance of classifiers in the transformed space, because locality in the original data space may not be kept in the transformed space. To address these problems, this paper presents a new approach for subclass division based on k-means clustering in the projected space, class by class using the iterative steps under EM-alike framework. Experiments are performed on the artificial data set, the UCI machine learning data sets, the CENPARMI handwritten numeral database, the NUST603 handwritten Chinese character database, and the terrain cover database. Extensive experimental results demonstrate the performance advantages of the proposed method.     

Heterogeneous forest classification by creating mixed vegetation classes using EO-1 Hyperion

Hyperspectral satellite data is an efficient tool in vegetation mapping; however, previous studies indicate that classifying heterogeneous forests might be difficult. In this study, we propose a mapping method for a heterogeneous forest using the data of the Earth Observing-1 (EO-1) Hyperion supplemented by field survey. We introduced a band reduction method to raise classification accuracy of the Support Vector Machine classification algorithm and compared the results to the one reduced by principal component analysis (PCA), stepwise discriminant analysis (SDA), and the original data set. We also used a modified version of the Vegetation–Impervious–Soil model to create mixed vegetation classes consisting of the commonly mixing species in the area and classified them using Decision Tree classification method. We managed to achieve 84.28% approximately using our band reduction method which is 2.36% increase compared to PCA (81.92%), 1.43% compared to the SDA (82.85%), and 7.61% compared to the original data set (76.67%). Introducing the mixed vegetation classes raised the overall accuracy even higher (85.79%).     

Quality grade identification of green tea using the eigenvalues of PCA based on the E-nose signals

The potential of the electronic nose to monitor Longjing tea different grade based on dry tea leaf, tea beverages and tea remains volatiles was studied. The original feature vector was obtained from the response signals of the E-nose, and was analyzed by principal component analysis (PCA). To decrease the data dimension and optimize the feature vector, the front five principal component values of the PCA were extracted as the final feature vectors by PCA. The linear discrimination analysis (LDA) and the back-propagation neural network (BPNN) were proposed to identify Longjing tea grade. The results showed that the discrimination results and testing results for the tea grade were better based on the tea beverages than those based on the tea leaf and the tea remains based on the new five feature vectors; both of the LDA and BPNN methods achieved better discrimination for the tea grades based on the tea beverages and the analysis results of the two methods were accordance.     

基于逐步判别分析和BP神经网络的电子鼻猪肉储藏时间预测

旨在探讨一种快速检测猪肉储藏时间的电子鼻方法.本研究采用德国Airsense公司的PEN 2型便携式电子鼻对不同储藏时间(0～7 d)的猪肉样品进行检测,每天检测42个样品,每个样品质量为10 g,密封时间为5 min.提取第60 s数据进行线性判别分析,结果显示电子鼻能较好的区分不同储藏天数的猪肉样品.同时用逐步判别分析和BP神经网络对猪肉储藏时间进行预测,训练集的准确率,前者为100%,后者为94.17%,而预测集的准确率,前者为97.92%,后者为93.75%.研究表明电子鼻技术有望在猪肉新鲜度快速检测上得到广泛的应用.     

Discrimination of green tea quality using the electronic nose technique and the human panel test, comparison of linear and nonlinear classification tools

Electronic nose (E-nose) technique was attempted to discriminate green tea quality instead of human panel test in this work. Four grades of green tea, which were classified by the human panel test, were attempted in the experiment. First, the E-nose system with eight metal oxide semiconductors gas sensors array was developed for data acquisition; then, the characteristic variables were extracted from the responses of the sensors; next, the principal components (PCs), as the input of the discrimination model, were extracted by principal component analysis (PCA); finally, three different linear or nonlinear classification tools, which were K-nearest neighbors (KNN), artificial neural network (ANN) and support vector machine (SVM), were compared in developing the discrimination model. The number of PCs and other model parameters were optimized by cross-validation. Experimental results showed that the performance of SVM model was superior to other models. The optimum SVM model was achieved when 4 PCs were included. The back discrimination rate was equal to 100% in the training set, and predictive discrimination rate was equal to 95% in the prediction set, respectively. The overall results demonstrated that E-nose technique with SVM classification tool could be successfully used in discrimination of green tea's quality, and SVM algorithm shows its superiority in solution to classification of green tea's quality using E-nose data.     

最坏分离的联合分辨率判别分析

现实中,常需辨识低分辨率(low-resolution,简称LR)图像(如监控系统所捕捉的人脸),但相比通常的高(high-resolution,简称HR)或超(super-resolution,简称SR)分辨率图像而言,其含有相对较少的判别信息,致使通常的子空间学习算法,如结合主成分分析(principal components analysis,简称PCA)的线性判别分析(linear discriminant analysis,简称LDA)难以获得理想的识别效果.为了缓和该问题,最近所提出的联合判别分析(如SDA)借助与低分辨率相配对的高分辨率图像辅助设计LR图像分类器.在SDA的实现中,其采用了类似LDA的平均散度定义,使SDA遗传了LDA在投影时难以使相对靠近的类充分分离的问题.为了克服该不足,提出了针对LR图像识别的最坏分离的联合分辨率判别分析(worst-separated couple-resolution discriminant analysis,简称WSCR),从而使:(1) LR和HR投影到同一低维子空间;(2) 投影后的最小类间隔最大化.实验结果表明:与SDA相比,WSCR更适用于低分辨率的图像识别.     

A support vector regression based prediction model of affective responses for product form design

In this paper, a state-of-the-art machine learning approach known as support vector regression (SVR) is introduced to develop a model that predicts consumers’ affective responses (CARs) for product form design. First, pairwise adjectives were used to describe the CARs toward product samples. Second, the product form features (PFFs) were examined systematically and then stored them either as continuous or discrete attributes. The adjective evaluation data of consumers were gathered from questionnaires. Finally, prediction models based on different adjectives were constructed using SVR, which trained a series of PFFs and the average CAR rating of all the respondents. The real-coded genetic algorithm (RCGA) was used to determine the optimal training parameters of SVR. The predictive performance of the SVR with RCGA (SVR–RCGA) is compared to that of SVR with 5-fold cross-validation (SVR–5FCV) and a back-propagation neural network (BPNN) with 5-fold cross-validation (BPNN–5FCV). The experimental results using the data sets on mobile phones and electronic scooters show that SVR performs better than BPNN. Moreover, the RCGA for optimizing training parameters for SVR is more convenient for practical usage in product form design than the timeconsuming CV.     

Rapid differentiation between E. coli and Salmonella Typhimurium using metal oxide sensors integrated with pattern recognition

A rapid method to differentiate between E coli and Salmonella Typhimurium was developed. E. coli and S. Typhimurium were separately grown in super broth and incubated at 37 °C. Super broth without inoculation of E. coli or S. Typhimurium was used as control. Numbers of E. coli and S. Typhimurium were followed using a colony counting method. Identification of the volatile metabolites produced by E. coli and S. Typhimurium was determined using solid-phase microextraction coupled with gas chromatography/mass spectrometry. An electronic nose with 12 non-specific metal oxide sensors was used to monitor the volatile profiles produced by E. coli and S. Typhimurium. Principal component analysis (PCA) and back-propagation neural network (BPNN) were used as pattern recognition tools. PCA was used for data exploration and dimensional reduction. PCA could visualize class separation between sample subgroups. The BPNN was shown to be capable of predicting the number of E. coli and S. Typhimurium. Good prediction was possible as measured by a regression coefficient (R² = 0.96) between true and predicted data. Using metal oxide sensors and pattern recognition techniques, it was possible to discriminate between samples containing E. coli from those containing S. Typhimurium.     

堆叠去噪自编码器在垃圾邮件过滤中的应用

针对垃圾邮件数量日益攀升的问题,提出了将堆叠去噪自编码器应用到垃圾邮件分类中.首先,在无标签数据集上,使用无监督学习方法最小化重构误差,对堆叠去噪自编码器进行贪心逐层预训练,从而获得原始数据更加抽象和健壮的特征表示; 然后,在堆叠去噪自编码器的最上层添加一个分类器后,在有标签数据集上,利用有监督学习方法最小化分类误差,对预训练获得的网络参数进行微调,获得最优化的模型; 最后, 利用训练完成的堆叠去噪编码器在6个不同的公开数据集上进行测试.将准确率、召回率、更具有平衡性的马修斯相关系数作为实验性能评价标准,实验结果表明,相比支持向量机算法、贝叶斯方法和深度置信网络的分类效果,基于堆叠去噪自编码器的垃圾邮件分类器的准确率都高于95%,马修斯相关系数都大于0.88,在应用中具有更高的准确率和更好的健壮性.     

Design and Construction of Electronic Nose for Multi-purpose Applications by Sensor Array Arrangement Using IBGSA

Technological progresses in the gas sensor fields provide the possibility of designing and construction of Electronic nose (E-nose) based on the Biological nose. E-nose uses specific hardware and software units; Sensor array is one of the critical units in the E-nose and its types of sensors are determined based on the application. So far, many achievements have been reported for using the E-nose in different fields of application. In this work, an E-nose for handling multi-purpose applications is proposed, and the employed hardware and pattern recognition techniques are depicted. To achieve higher recognition rate and lower power consumption, the improved binary gravitational search algorithm (IBGSA) and the K-nearest neighbor (KNN) classifier are used for automatic selecting the best combination of the sensors. The designed E-nose is tested by classifying the odors in different case studies, including moldy bread recognition in food and beverage field, herbs recognition in the medical field, and petroleum products recognition in the industrial field. Experimental results confirm the efficiency of the proposed method for E-nose realization.     

环境化合物降解预测模型的研究

为建立化合物降解的计算机预测模型,确定降解和非降解化合物显然不同的参数.选择389个有机分子作为数据集,选其中312个为训练集,其余77个为验证集,每个分子计算195个分子参数,分别采用逐步判别法和主成分分析法建模,并用外部验证集验证模型的预测能力.结果:逐步判别法分析结果中,训练集的降解和非降解化合物的正确率分别为90.6%和69.5%;验证集的降解和非降解化合物的正确率分别为83.9%和63.6%.主成分分析结果在测试集中,降解和非降解化合物的正确率分别为80.4%和31.8%.验证集的降解化合物和非降解化合物的正确率分别为67.9%和50.0%.因此,采用逐步判别法模拟而建立的数学模型,可作为预测化合物降解的模型.以卜研究可以为预测有机物降解提供参考.     

电子鼻判别小麦陈化年限的检测方法研究

采用电子鼻对五个储藏年限的陈化小麦进行年限分析,确定了采用电子鼻判别小麦储藏年限的最佳参数及方法.对传感器信号进行多因素方差分析可知:对于固定容器的陈化小麦样品,不同的小麦密封时间对电子鼻的响应信号的影响极为显著;其次是小麦在烧杯内的密封质量.通过静置密封时间和密封质量的方差分析,得出小麦在500 mL烧杯内的最佳静置时间为1.5 h,密封在烧杯内的小麦最恰当质量为50 g.采用以上参数,对五个储藏年限的小麦进行辨别,PCA分析可以将不同储藏年限的小麦较好的区分开来,并且五个年份的小麦自右上角至左下角依次排列;而LDA分析能够将差别年限较大的陈化小麦进行区分,差距较小的,不能够很好的区分,其区分效果不如PCA分析;进而采用BP神经网络的方法进行判别分析,训练样本正确率为100%,测试样本正确率也达到了85%以上.     

Segmented canonical discriminant analysis of in situ hyperspectral data for identifying 13 urban tree species

A total of 458 in situ hyperspectral data were collected from 13 urban tree species in the City of Tampa, FL, USA using a spectrometer. The 13 species include 11 broadleaf and two conifer species. Three different techniques, segmented canonical discriminant analysis (CDA), segmented principal component analysis (PCA) and segmented stepwise discriminate analysis (SDA), were applied and compared for dimension reduction and feature extraction. With each of the three techniques, 10 features were extracted or selected from four spectral regions, visible (VIS: 1412–1797 nm), near-infrared (NIR: 707–1352 nm), mid-infrared 1 (MIR1: 1412–1797 nm) and mid-infrared 2 (MIR2: 1942–2400 nm), and used to discriminate the 13 urban tree species with a linear discriminate analysis (LDA) method. The cross-validation results, based on training samples that were used in the feature reduction step, and the results calculated from the test samples were used for evaluating the ability of the in situ hyperspectral data and performance of the segmented CDA, PCA and SDA to identify the 13 tree species. The experimental results indicate that a satisfactory discrimination of the 13 tree species was achieved using the segmented CDA technique (average accuracy (AA) = 96%, overall accuracy (OAA) = 96% and kappa = 0.958 from the cross-validation results; AA = 90%, OAA = 90% and kappa = 0.896 from the test samples) compared to the segmented PCA and SDA techniques, respectively (AA = 76% and 86%, OAA = 78% and 87%, and kappa = 0.763 and 0.857 from the cross-validation results; AA = 79% and 88%, OAA = 80% and 89%, and kappa = 0.782 and 0.879 from the test samples). In this study, the segmented CDA transformation is effective for dimension reduction and feature extraction for species discrimination with a relatively limited number of training samples. It outperformed the segmented PCA and SDA methods and produced the highest accuracies. The NIR and MIR1 regions have greater power for identifying the 13 species compared to the VIS and MIR2 spectral regions. The results indicate that CDA or segmented CDA could be applied broadly in mapping forest cover types, species identification and/or other land use/land cover classification practices with hyperspectral remote sensing data.     

Patterns of irrigated rice growth and malaria vector breeding in Mali using multi-temporal ERS-2 synthetic aperture radar

We explored the use of the European Remote Sensing Satellite 2 Synthetic Aperture Radar (ERS-2 SAR) to trace the development of rice plants in an irrigated area near Niono, Mali and relate that to the density of anopheline mosquitoes, especially An. gambiae. This is important because such mosquitoes are the major vectors of malaria in sub-Saharan Africa, and their development is often coupled to the cycle of rice development. We collected larval samples, mapped rice fields using GPS and recorded rice growth stages simultaneously with eight ERS-2 SAR acquisitions. We were able to discriminate among rice growth stages using ERS-2 SAR backscatter data, especially among the early stages of rice growth, which produce the largest numbers of larvae. We could also distinguish between basins that produced high and low numbers of anophelines within the stage of peak production. After the peak, larval numbers dropped as rice plants grew taller and thicker, reducing the amount of light reaching the water surface. ERS-2 SAR backscatter increased concomitantly. Our data support the belief that ERS-2 SAR data may be helpful for mapping the spatial patterns of rice growth, distinguishing different agricultural practices, and monitoring the abundance of vectors in nearby villages.     

A novel Bayesian logistic discriminant model: An application to face recognition

The linear discriminant analysis (LDA) is a linear classifier which has proven to be powerful and competitive compared to the main state-of-the-art classifiers. However, the LDA algorithm assumes the sample vectors of each class are generated from underlying multivariate normal distributions of common covariance matrix with different means (i.e., homoscedastic data). This assumption has restricted the use of LDA considerably. Over the years, authors have defined several extensions to the basic formulation of LDA. One such method is the heteroscedastic LDA (HLDA) which is proposed to address the heteroscedasticity problem. Another method is the nonparametric DA (NDA) where the normality assumption is relaxed. In this paper, we propose a novel Bayesian logistic discriminant (BLD) model which can address both normality and heteroscedasticity problems. The normality assumption is relaxed by approximating the underlying distribution of each class with a mixture of Gaussians. Hence, the proposed BLD provides more flexibility and better classification performances than the LDA, HLDA and NDA. A subclass and multinomial versions of the BLD are proposed. The posterior distribution of the BLD model is elegantly approximated by a tractable Gaussian form using variational transformation and Jensen's inequality, allowing a straightforward computation of the weights. An extensive comparison of the BLD to the LDA, support vector machine (SVM), HLDA, NDA and subclass discriminant analysis (SDA), performed on artificial and real data sets, has shown the advantages and superiority of our proposed method. In particular, the experiments on face recognition have clearly shown a significant improvement of the proposed BLD over the LDA.     

Electronic nose technique potential monitoring mandarin maturity

Over the past years, electronic nose technology opened the possibility to exploit information on behavior aroma to assess fruit ripening stage. The objective in this study was to evaluate the capacity of electronic nose to monitoring the change in volatile production of mandarin during different picking-date, using a specific electronic nose device (PEN 2). Principal component analysis (PCA) and linear discriminant analysis (LDA) were used in order to investigate whether the electronic nose was able to distinguish among different picking-date (ripeness states). The loadings analysis was used to identify the sensors responsible for discrimination in the current pattern file. The results obtained prove that the electronic nose PEN 2 can discriminate successfully different picking-date on mandarin using LDA analysis. But, electronic nose was not able to detect a clear difference in volatile profile on mandarin using PCA analysis. During external validation using LDA was obtained to classified 92% of the total samples properly. Some sensors have the highest influence in the current pattern file for electronic nose PEN 2. A subset of few sensors can be chosen to explain all the variance. This result could be used in further studies to optimize the number of sensors.     

Text categorization algorithms using semantic approaches, corpus-based thesaurus and WordNet

In this paper, a corpus-based thesaurus and WordNet were used to improve text categorization performance. We employed the k-NN algorithm and the back propagation neural network (BPNN) algorithms as the classifiers. The k-NN is a simple and famous approach for categorization, and the BPNNs has been widely used in the categorization and pattern recognition fields. However the standard BPNN has some generally acknowledged limitations, such as a slow training speed and can be easily trapped into a local minimum. To alleviate the problems of the standard BPNN, two modified versions, Morbidity neurons Rectified BPNN (MRBP) and Learning Phase Evaluation BPNN (LPEBP), were considered and applied to the text categorization. We conducted the experiments on both the standard reuter-21578 data set and the 20 Newsgroups data set. Experimental results showed that our proposed methods achieved high categorization effectiveness as measured by the precision, recall and F-measure protocols.     

Application of Kernel-Based Feature Space Transformations and Learning Methods to Phoneme Classification

This paper examines the applicability of some learning techniques to the classification of phonemes. The methods tested were artificial neural nets (ANN), support vector machines (SVM) and Gaussian mixture modeling (GMM). We compare these methods with a traditional hidden Markov phoneme model (HMM), working with the linear prediction-based cepstral coefficient features (LPCC). We also tried to combine the learners with linear/nonlinear and unsupervised/supervised feature space transformation methods such as principal component analysis (PCA), independent component analysis (ICA), linear discriminant analysis (LDA), springy discriminant analysis (SDA) and their nonlinear kernel-based counterparts. We found that the discriminative learners can attain the efficiency of HMM, and that after the transformations they can retain the same performance in spite of the severe dimension reduction. The kernel-based transformations brought only marginal improvements compared to their linear counterparts.