Estimating canopy closure density and above-ground tree biomass using partial least square methods in Chinese boreal forests

Boreal forests play an important role in global environment systems. Understanding boreal forest ecosystem structure and function requires accurate monitoring and estimating of forest canopy and biomass. We used partial least square regression （PLSR） models to relate forest parameters, i.e. canopy closure density and above ground tree biomass, to Landsat ETM＋ data. The established models were optimized according to the variable importance for projection （VIP） criterion and the bootstrap method, and their performance was compared using several statistical indices. All variables selected by the VIP criterion passed the bootstrap test （p〈0.05）. The simplified models without insignificant variables （VIP 〈1） performed as well as the full model but with less computation time. The relative root mean square error （RMSE%） was 29% for canopy closure density, and 58% for above ground tree biomass. We conclude that PLSR can be an effective method for estimating canopy closure density and above ground biomass.

Factors influencing the internet banking adoption decision in North Cyprus: an evidence from the partial least square approach of the structural equation modeling

Purpose:This paper aims to examine how the adoption decision of the internet banking in North Cyprus would be affected based on the following dimensions;the technology features,the personal characteristics,the social environment and the expected risk.Design/methodology/approach:A self-administered survey was conducted with 291 participants responded to it.The partial least square approach of the structural equation modeling(PLS-SEM)is employed to investigate the direct effects of the proposed factors on the adoption decision.Additionally,the mediation test is used to examine indirect effects.Findings:Results showed that even though the participants appreciated the benefits of the online banking as the perceived usefulness factor exerts the greatest direct effect,they would rather use clear and easy-to-use websites,adding to that their assessments of the usefulness of these services are significantly influenced by the surrounding people’s views and prior experience.This is demonstrated by the total effects of the perceived ease of use and the subjective norm factors,which are greater than the direct effect of the perceived usefulness factor since both of these factors have significant direct and indirect effects mediated by the perceived usefulness factor.The negative impact of the perceived risk factor is weak compared to the previous factors.While the personal innovativeness factor showed the weakest effect among the proposed factors.

Partial least square modeling of hydrolysis: analyzing the impacts of pH and acetate

pH and volatile fatty acids both might affect the further hydrolysis of particulate solid waste, which is the limiting-step of anaerobic digestion. To clarify the individual effects of pH and volatile fatty acids, batch experiments were conducted at fixed pH value （pH 5-9） with or without acetate （20 g/L）. The hydrolysis efficiencies of carbohydrate and protein were evaluated by carbon and nitrogen content of solids, amylase activity and proteinase activity. The trend of carbohydrate hydrolysis with pH was not affected by the addition of acetate, following the sequence ofpH 7〉pH 8〉pH 9〉pH 6〉pH 5; but the inhibition of acetate （20 g/L） was obvious by 10%-60 %. The evolution of residual nitrogen showed that the effect of pH on protein hydrolysis was minor, while the acetate was seriously inhibitory especially at alkali condition by 45%-100 %. The relationship between the factors （pH and acetate） and the response variables was evaluated by partial least square modeling （PLS）. The PLS analysis demonstrated that the hydrolysis of carbohydrate was both affected by pH and acetate, with pH the more important factor. Therefore, the inhibition by acetate on carbohydrate hydrolysis was mainly due to the corresponding decline of pH, but the presence of acetate species, while the acetate species was the absolutely important factor for the hydrolysis of protein.

DATA MODELING METHOD BASED ON PARTIAL LEAST SQUARE REGRESSION AND APPLICATIO N IN CORRELATION ANALYSIS OF THE STATOR BARS CONDITION PARAMETERS

Objective To investigate v arious data message of the stator bars condition parameters under the condition that only a few samples are available, especially about correlation information between the nondestructive parameters and residual breakdown voltage of the stat or bars. Methods Artificial stator bars is designed to simulat e the generator bars. The partial didcharge( PD) and dielectric loss experiments are performed in order to obtain the nondestructive parameters, and the residua l breakdown voltage acquired by AC damage experiment. In order to eliminate the dimension effect on measurement data, raw data is preprocessed by centered-compr ess. Based on the idea of extracting principal components, a partial least squar e (PLS) method is applied to screen and synthesize correlation information betwe en the nondestructive parameters and residual breakdown voltage easily. Moreover , various data message about condition parameters are also discussed. Re sults Graphical analysis function of PLS is easily to understand vario us data message of the stator bars condition parameters. The analysis Results ar e consistent with result of aging testing. Conclusion The meth od can select and extract PLS components of condition parameters from sample dat a, and the problems of less samples and multicollinearity are solved effectively in regression analysis.

Characteristic wavelength selection of volatile organic compounds infrared spectra based on improved interval partial least squares

As important components of air pollutant,volatile organic compounds(VOCs)can cause great harm to environment and human body.The concentration change of VOCs should be focused on in real-time environment monitoring system.In order to solve the problem of wavelength redundancy in full spectrum partial least squares(PLS)modeling for VOCs concentration analysis,a new method based on improved interval PLS(iPLS)integrated with Monte-Carlo sampling,called iPLS-MC method,was proposed to select optimal characteristic wavelengths of VOCs spectra.This method uses iPLS modeling to preselect the characteristic wavebands of the spectra and generates random wavelength combinations from the selected wavebands by Monte-Carlo sampling.The wavelength combination with the best prediction result in regression model is selected as the characteristic wavelengths of the spectrum.Different wavelength selection methods were built,respectively,on Fourier transform infrared(FTIR)spectra of ethylene and ethanol gas at different concentrations obtained in the laboratory.When the interval number of iPLS model is set to 30 and the Monte-Carlo sampling runs 1000 times,the characteristic wavelengths selected by iPLS-MC method can reduce from 8916 to 10,which occupies only 0.22%of the full spectrum wavelengths.While the RMSECV and correlation coefficient(Rc)for ethylene are 0.2977 and 0.9999 ppm,and those for ethanol gas are 0.2977 ppm and 0.9999.The experimental results show that the iPLS-MC method can select the optimal characteristic wavelengths of VOCs FTIR spectra stably and effectively,and the prediction performance of the regression model can be significantly improved and simplified by using characteristic wavelengths.

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.

Model of Hot Metal Silicon Content in Blast Furnace Based on Principal Component Analysis Application and Partial Least Square

In blast furnace （BF） iron-making process, the hot metal silicon content was usually used to measure the quality of hot metal and to reflect the thermal state of BF. Principal component analysis （PCA） and partial least- square （PLS） regression methods were used to predict the hot metal silicon content. Under the conditions of BF rela- tively stable situation, PCA and PLS regression models of hot metal silicon content utilizing data from Baotou Steel No. 6 BF were established, which provided the accuracy of 88.4% and 89.2%. PLS model used less variables and time than principal component analysis model, and it was simple to calculate. It is shown that the model gives good results and is helpful for practical production.

Process Characterization of the Transesterification of Rapeseed Oil to Biodiesel Using Design of Experiments and Infrared Spectroscopy

For optimization of production processes and product quality,often knowledge of the factors influencing the process outcome is compulsory.Thus,process analytical technology(PAT)that allows deeper insight into the process and results in a mathematical description of the process behavior as a simple function based on the most important process factors can help to achieve higher production efficiency and quality.The present study aims at characterizing a well-known industrial process,the transesterification reaction of rapeseed oil with methanol to produce fatty acid methyl esters(FAME)for usage as biodiesel in a continuous micro reactor set-up.To this end,a design of experiment approach is applied,where the effects of two process factors,the molar ratio and the total flow rate of the reactants,are investigated.The optimized process target response is the FAME mass fraction in the purified nonpolar phase of the product as a measure of reaction yield.The quantification is performed using attenuated total reflection infrared spectroscopy in combination with partial least squares regression.The data retrieved during the conduction of the DoE experimental plan were used for statistical analysis.A non-linear model indicating a synergistic interaction between the studied factors describes the reactor behavior with a high coefficient of determination(R^(2))of 0.9608.Thus,we applied a PAT approach to generate further insight into this established industrial process.

Comparison of Artificial Neural Networks with Partial Least Squares Regression for Simultaneous Determinations by ICP-AES

Simultaneous determination of several elements （U, Ta, Mn, Zr and W） with inductively coupled plasma atomic emission spectrometry （ICP-AES） in the presence of spectral interference was performed using chemometrics methods. True comparison between artificial neural network （ANN） and partial least squares regression （PLS） for simultaneous determination in different degrees of overlap was investigated. The emission spectra were recorded at uranium analytical line （263.553 nm） with a 0.06 nm spectral window by ICP-AES. Principal component analysis was applied to data and scores on 5 dominant principal components were subjected to ANN. A 5-5-5 （input, hidden and output neurons） network was used with linear transfer function after both hidden and output layers. The PI,S model was trained with five latent variables and 20 samples in calibration set. The relative errors of predictions （REP） in test set were 3.75% and 3.56% for ANN and PLS respectively.

A Comparison of CNN and PLSR for Glucose Monitoring Using Mid-Infrared Absorption Spectroscopy

With the development of mid-infrared (MIR) photoelectric devices, mid-infrared spectroscopy has become one of the important methods for non-invasive detection of blood glucose. The mid-infrared region (4000 - 400 cm-1) has the well-known fingerprint region (1200 - 800 cm-1) of glucose, which has clearer characteristic absorption peaks and better specificity. There is a lot of molecular information about glucose in the MIR. The non-invasive detection of blood glucose by mid-infrared spectroscopy needs to achieve certain accuracy, and the quantitative model is an important factor affecting the accuracy of glucose detection. In this paper, the samples of imitation solution containing only glucose and the samples of imitation mixed solution are taken as the research objects, and the mid-infrared spectral data of the samples are collected. The full spectrum partial least squares Regression (PLSR) model, SNV + Ctr-PLSR model, MSC + Ctr-PLSR model, and convolutional neural networks (CNN) model of 3000 - 900 cm-1 band were constructed. Full spectrum PLS model and CNN model of 1200 - 900 cm-1 band were constructed. The experimental results show that the optimal model of the two bands is CNN, then the correlation coefficient of prediction set (Rp) of 3000 - 900 cm-1 band is 0.95, and the root mean square error of pre-diction set (RMSEP) value is 22.10. The Rp of 1200 - 900 cm-1 band is 0.95, and the RMSEP value is 22.54. The research results show that CNN is a promising method, which has higher accuracy than PLSR, and is especially suitable for modeling human complex environment. In addition, the study provides a theoretical and practical basis for CNN in feature selection and model interpretation.

Spectroscopic Leaf Level Detection of Powdery Mildew for Winter Wheat Using Continuous Wavelet Analysis

Powdery mildew （Blumeria graminis） is one of the most destructive crop diseases infecting winter wheat plants, and has devastated millions of hectares of farmlands in China. The objective of this study is to detect the disease damage of powdery mildew on leaf level by means of the hyperspectral measurements, particularly using the continuous wavelet analysis. In May 2010, the reflectance spectra and the biochemical properties were measured for 114 leaf samples with various disease severity degrees. A hyperspectral imaging system was also employed for obtaining detailed hyperspectral information of the normal and the pustule areas within one diseased leaf. Based on these spectra data, a continuous wavelet analysis （CWA） was carried out in conjunction with a correlation analysis, which generated a so-called correlation scalogram that summarizes the correlations between disease severity and the wavelet power at different wavelengths and decomposition scales. By using a thresholding approach, seven wavelet features were isolated for developing models in determining disease severity. In addition, 22 conventional spectral features （SFs） were also tested and compared with wavelet features for their efficiency in estimating disease severity. The multivariate linear regression （MLR） analysis and the partial least square regression （PLSR） analysis were adopted as training methods in model mildew on leaf level were found to be closely related with the development. The spectral characteristics of the powdery spectral characteristics of the pustule area and the content of chlorophyll. The wavelet features performed better than the conventional SFs in capturing this spectral change. Moreover, the regression model composed by seven wavelet features outperformed （R2=0.77, relative root mean square error RRMSE=0.28） the model composed by 14 optimal conventional SFs （R2---0.69, RRMSE--0.32） in estimating the disease severity. The PLSR method yielded a higher accuracy than the MLR method. A combination of CWA and PLSR was found to be promising in providing relatively accurate estimates of disease severity of powdery mildew on leaf level.

Application of Wavelet Transform in the Prediction of Navel Orange Vitamin C Content by Near-Infrared Spectroscopy

This study was to search for an approach for rapid measurement of orange vitamin C （Vc） content. By using different decomposing levels of Daubechies 3 wavelet transform, the near-infrared spectra signals obtained from intact fruits of 100 navel orange samples were denoised, and the results of the predicted Vc contents for the corresponding samples determined by the reconstructed spectra after denoising were validated by means of PLS-CV （partial least squared-cross validation）. It was shown that the prediction effects verified by PLS-CV analysis varied when different wavelet transform decomposing levels were employed. At the wavelet decomposing level 4, the best prediction effect was obtained, with the correlation coefficient R between the prediction and true values being 0.9574 and the expected variance RMSECV being as low as 3.9 mg 100 g^-1. Furthermore, the 11 different approaches for the pretreatment of the near-infrared spectrum were compared. It was found that the calibration model established by PLS using spectra pretreated by wavelet transform denoising provided the best prediction for Vc content, exhibiting the highest correlation between the prediction and true values by cross validation. In conclusion, the near infrared spectral model denoised by means of wavelet transform can be used for accurate, rapid, and nondestructive quantitative analysis on navel orange Vc content.

Application of Near Infrared Diffuse Reflectance Spectroscopy with Radial Basis Function Neural Network to Determination of Rifampincin Isoniazid and Pyrazinamide Tablets

Partial least squares（PLS）,back-propagation neural network（BPNN）and radial basis function neural network（RBFNN）were respectively used for estalishing quantative analysis models with near infrared（NIR）diffuse reflectance spectra for determining the contents of rifampincin（RMP）,isoniazid（INH）and pyrazinamide（PZA）in rifampicin isoniazid and pyrazinamide tablets.Savitzky-Golay smoothing,first derivative,second derivative,fast Fourier transform（FFT）and standard normal variate（SNV）transformation methods were applied to pretreating raw NIR diffuse reflectance spectra.The raw and pretreated spectra were divided into several regions,depending on the average spectrum and RSD spectrum.Principal component analysis（PCA）method was used for analyzing the raw and pretreated spectra in different regions in order to reduce the dimensions of input data.The optimum spectral regions and the models＇ parameters were chosen by comparing the root mean square error of cross-validation（RMSECV）values which were obtained by leave-one-out cross-validation method.The RMSECV values of the RBFNN models for determining the contents of RMP,INH and PZA were 0.00288,0.00226 and 0.00341,respectively.Using these models for predicting the contents of INH,RMP and PZA in prediction set,the RMSEP values were 0.00266,0.00227 and 0.00411,respectively.These results are better than those obtained from PLS models and BPNN models.With additional advantages of fast calculation speed and less dependence on the initial conditions,RBFNN is a suitable tool to model complex systems.

New Descriptors of Amino Acids and Its Applications to Peptide Quantitative Structure-activity Relationship

A new set of descriptors, HSEHPCSV （component score vector of hydrophobic, steric, and electronic properties together with hydrogen bonding contributions）, were derived from principal component analyses of 95 physicochemical variables of 20 natural amino acids separately according to different kinds of properties described, namely, hydrophobic, steric, and electronic properties as well as hydrogen bonding contributions. HSEHPCSV scales were then employed to express structures of angiotensin-converting enzyme inhibitors, bitter tasting thresholds and bactericidal 18 peptide, and to construct QSAR models based on partial least square （PLS）. The results obtained are as follows： the multiple correlation coefficient （R2cum） of 0.846, 0.917 and 0.993, leave-one-out cross validated Q2cm of 0.835, 0.865 and 0.899, and root-mean-square error for estimated error （RMSEE） of 0.396, 0.187and 0.22, respectively. Satisfactory results showed that, as new amino acid scales, data of HSEHPCSV may be a useful structural expression methodology＇for the studies on peptide QSAR （quantitative structure-activity relationship） due to many advantages such as plentiful structural information, definite physical and chemical meaning and easy interpretation.

Using Support Vector Machine to Predict Eco-environment Burden:A Case Study of Wuhan,Hubei Province,China

Objective The human socio-economic development depends on the planet＇s natural capital. Humans have had a considerable impact on the earth, such as resources depression and environment deterioration. The objective of this study was to assess the impact of socio-economic development on the ecological environment of Wuhan, Hubei Province, China, during the general planning period 2006-2020. Methods Support vector machine （SVM） model was constructed to simulate the process of eco-economic system of Wuhan. Socio-economic factors of urban total ecological footprint （TEF） were selected by partial least squares （PLS） and leave-one-out cross validation （LOOCV）. Historical data of socio-economic factors as inputs, and corresponding historical data of TEF as target outputs, were presented to identify and validate the SVM model. When predicted input data after 2005 were presented to trained model as generalization sets, TEFs of 2005, 2006,…, till 2020 were simulated as output in succession. Results Up to 2020, the district would have suffered an accumulative TEF of 28.374 million gha, which was over 1.5 times that of 2004 and nearly 3 times that of 1988. The per capita EF would be up to 3.019 gha in 2020. Contusions The simulation indicated that although the increase rate of GDP would be restricted in a lower level during the general planning period, urban ecological environment burden could not respond to the socio-economic circumstances promptly. SVM provides tools for dynamic assessment of regional eco-environment. However, there still exist limitations and disadvantages in the model. We believe that the next logical step in deriving better dynamic models of ecosystem is to integrate SVM and other algorithms or technologies.

Enhancing the Effectiveness of Trimethylchlorosilane Purification Process Monitoring with Variational Autoencoder

In modern industry,process monitoring plays a significant role in improving the quality of process conduct.With the higher dimensional of the industrial data,the monitoring methods based on the latent variables have been widely applied in order to decrease the wasting of the industrial database.Nevertheless,these latent variables do not usually follow the Gaussian distribution and thus perform unsuitable when applying some statistics indices,especially the T^(2) on them.Variational AutoEncoders(VAE),an unsupervised deep learning algorithm using the hierarchy study method,has the ability to make the latent variables follow the Gaussian distribution.The partial least squares(PLS)are used to obtain the information between the dependent variables and independent variables.In this paper,we will integrate these two methods and make a comparison with other methods.The superiority of this proposed method will be verified by the simulation and the Trimethylchlorosilane purification process in terms of the multivariate control charts.

Comprehensive Evaluation and Prediction of the Effectiveness of H_(2)O_(2)- assisted Na_(2)CO_(3)Pretreatment of Corn Stover Using Multivariate Analysis

In this study,multivariate analysis methods,including a principal component analysis(PCA)and partial least square(PLS)analysis,were applied to reveal the inner relationship of the key variables in the process of H_(2)O_(2)-assisted Na_(2)CO_(3)(HSC)pretreatment of corn stover.A total of 120 pretreatment experiments were implemented at the lab scale under different conditions by varying the particle size of the corn stover and process variables.The results showed that the Na_(2)CO_(3) dosage and pretreatment temperature had a strong influence on lignin removal,whereas pulp refining instrument(PFI)refining and Na_(2)CO_(3) dosage played positive roles in the final total sugar yield.Furthermore,it was found that pretreatment conditions had a more significant impact on the amelioration of pretreatment effectiveness compared with the properties of raw corn stover.In addition,a prediction of the effectiveness of the corn stover HSC pretreatment based on a PLS analysis was conducted for the first time,and the test results of the predictability based on additional pretreatment experiments proved that the developed PLS model achieved a good predictive performance(particularly for the final total sugar yield),indicating that the developed PLS model can be used to predict the effectiveness of HSC pretreatment.Therefore,multivariate analysis can be potentially used to monitor and control the pretreatment process in future large-scale biorefinery applications.

Generalized Procrustes Analysis and External Preference Map Used to Consumer Drivers of Diet Gluten Free Product

In order to study correlations between sensory properties and acceptance, regular and gluten-free carrot cakes (sweetened with sucrose and sucralose) were evaluated. Appearance, aroma, flavor, texture and overall liking were analyzed by 120 carrot cake consumers using a 9-cm hedonic scale. Quantitative Descriptive Analysis (QDA) was carried out with 11 assessors among 16 attributes. Data were analyzed by ANOVA, Tukey test (p < 0.05), Internal Preference Map and Cluster analysis. Also texture parameters were analyzed by Partial Least Square (PLS) to be able to show the instrumental parameters influence. Generalized Procrustes Analysis (GPA) was used before an Ex-ternal Preference Mapping to reduce the scale effects and to obtain a consensus configuration. According to PLS correlation, the attributes hardness, fracturability, adhesiveness, chewiness and gumminess interfered on characteristics considered undesirable for texture acceptance and smoothness, elasticity, cohesiveness and water activity interfered as desirable features. It was noted that all cakes were well accepted, except cake sweetened with sucralose and mix done using cornmeal, rice flour, potato starch and corn starch (2:3:3:2). As for the instrumental aspects, cohesiveness and elasticity influenced positively the cake’s acceptance. Instead, smoothness and adhesion parameters weren’t so significant.

Quantitative analysis of ammonium salts in coking industrial liquid waste treatment process based on Raman spectroscopy

Quantitative analysis of ammonium salts in the process of coking industrial liquid waste treatment is successfully performed based on a compact Raman spectrometer combined with partial least square(PLS) method. Two main components(NH4SCN and(NH4)2S2O3) of the industrial mixture are investigated. During the data preprocessing, wavelet denoising and an internal standard normalization method are employed to improve the predicting ability of PLS models. Moreover,the PLS models with different characteristic bands for each component are studied to choose a best resolution. The internal and external calibration results of the validated model show a mass percentage error below 1% for both components.Finally, the repeatabilities and reproducibilities of Raman and reference titration measurements are also discussed.

The use of milk Fourier transform midinfrared spectra and milk yield to estimate heat production as a measure of efficiency of dairy cows

Background:Transformation of feed energy ingested by ruminants into milk is accompanied by energy losses via fecal and urine excretions,fermentation gases and heat.Heat production may differ among dairy cows despite comparable milk yield and body weight.Therefore,heat production can be considered an indicator of metabolic efficiency and directly measured in respiration chambers.The latter is an accurate but time-consuming technique.In contrast,milk Fourier transform mid-infrared(FTIR)spectroscopy is an inexpensive high-throughput method and used to estimate different physiological traits in cows.Thus,this study aimed to develop a heat production prediction model using heat production measurements in respiration chambers,milk FTIR spectra and milk yield measurements from dairy cows.Methods:Heat production was computed based on the animal’s consumed oxygen,and produced carbon dioxide and methane in respiration chambers.Heat production data included 16824-h-observations from 64 German Holstein and 20 dual-purpose Simmental cows.Animals were milked twice daily at 07:00 and 16:30 h in the respiration chambers.Milk yield was determined to predict heat production using a linear regression.Milk samples were collected from each milking and FTIR spectra were obtained with MilkoScan FT 6000.The average or milk yield-weighted average of the absorption spectra from the morning and afternoon milking were calculated to obtain a computed spectrum.A total of 288 wavenumbers per spectrum and the corresponding milk yield were used to develop the heat production model using partial least squares(PLS)regression.Results:Measured heat production of studied animals ranged between 712 and 1470 kJ/kg BW0.75.The coefficient of determination for the linear regression between milk yield and heat production was 0.46,whereas it was 0.23 for the FTIR spectra-based PLS model.The PLS prediction model using weighted average spectra and milk yield resulted in a cross-validation variance of 57%and a root mean square error of prediction of 86.5 kJ/kg BW0.75.The ratio of performance to deviation(RPD)was 1.56.Conclusion:The PLS model using weighted average FTIR spectra and milk yield has higher potential to predict heat production of dairy cows than models applying FTIR spectra or milk yield only.