A new issue of this journal has just
been published. To see abstracts of the papers it contains (with links through
to the full papers) click here:
Selected
papers from the latest issue:
Study of indole derivative inhibitors of Cytosolic phospholipase A2α based on Quantitative Structure Activity Relationship
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Xiaoquan Lu, Dongqin Ji, Jing Chen, Xibin Zhou, Haicai Shi
Cytosolic phospholipase A2α, one of the three subtypes of Cytosolic phospholipase A2 (α, β and γ), is deemed to play an important role in the arachidonate pathway. Due to the rate-limiting provider for proinflammatory mediators, it is a particularly attractive target for drug development. Studies have revealed that indol derivate compounds can inhibit the activities of Cytosolic phospholipase A2α. However, few papers on the relationship between the molecular structure and the activity of inhibitor were reported. In this study, the Quantitative Structure Activity Relationship (QSAR) of indole derivates has been performed based on the dataset of 49 compounds. By using stepwise multiple linear regression, 5 descriptors were selected from 1777 molecular descriptors, including GGI5 (Topological charge index G5), TIE(dssC) (sum of E-State of atom type dssC: ¦2S(dssC)), RDF115a (the atomic Sanderson ALOGP), RDF100c (the atomic charge), and RDF065p (the atomic polarizability). Subsequently, Partial Least Squares (PLS), Artificial Neural Networks (ANN) and Support Vector Machine (SVM) were adopted to build the QSAR model, respectively. The independent test indicated that the SVM can give the best statistical results. And indole derivative inhibitors activity might be related to global charge transfers, carbon atoms type linked benzyl sulfonamide and molecule geometrical the distance distribution.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Xiaoquan Lu, Dongqin Ji, Jing Chen, Xibin Zhou, Haicai Shi
Cytosolic phospholipase A2α, one of the three subtypes of Cytosolic phospholipase A2 (α, β and γ), is deemed to play an important role in the arachidonate pathway. Due to the rate-limiting provider for proinflammatory mediators, it is a particularly attractive target for drug development. Studies have revealed that indol derivate compounds can inhibit the activities of Cytosolic phospholipase A2α. However, few papers on the relationship between the molecular structure and the activity of inhibitor were reported. In this study, the Quantitative Structure Activity Relationship (QSAR) of indole derivates has been performed based on the dataset of 49 compounds. By using stepwise multiple linear regression, 5 descriptors were selected from 1777 molecular descriptors, including GGI5 (Topological charge index G5), TIE(dssC) (sum of E-State of atom type dssC: ¦2S(dssC)), RDF115a (the atomic Sanderson ALOGP), RDF100c (the atomic charge), and RDF065p (the atomic polarizability). Subsequently, Partial Least Squares (PLS), Artificial Neural Networks (ANN) and Support Vector Machine (SVM) were adopted to build the QSAR model, respectively. The independent test indicated that the SVM can give the best statistical results. And indole derivative inhibitors activity might be related to global charge transfers, carbon atoms type linked benzyl sulfonamide and molecule geometrical the distance distribution.
Filling and D-optimal designs for the correlated generalized exponential models
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
J.M. Rodríguez-Díaz, M.T. Santos-Martín, H. Waldl, M. Stehlík
The aim of this paper is to provide guidelines for the statistically efficient estimation of parameters of a modified Arrhenius model for chemical kinetics. A modified Arrhenius model is used for instance by modeling a flux of methane in troposphere or by chemical kinetics for reactions at membranes. D-optimal and filling designs for the Generalized Exponential Model with correlated observations are studied, considering the exponential covariance with or without nugget effect. Both equidistant and exact designs for small samples are examined, studying the behavior of different types of filling designs when a greater number of observations is preferred. Probably the main lesson we can learn is that the D-optimal design is analytically peculiar and these designs can be practically obtained only by numerical computation; however, specially two point locally D-optimal designs are very interesting, since they may help us to find a reasonable range for filling designs. The latter ones are probably only applicable when seeking for a higher number of design points. It is an interesting issue that very often the best designs do not use the whole design interval, but only a part of it; this should be taken into account by practitioners when they design their experiments. The second important observation is the large bias of the ML estimator of the correlation parameter. From the theoretical point of view this is not surprising since variance and correlation parameters are not simultaneously identifiable. We develop a bias reduction method and illustrate its effectiveness. We also provide practical implications for chemometrics.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
J.M. Rodríguez-Díaz, M.T. Santos-Martín, H. Waldl, M. Stehlík
The aim of this paper is to provide guidelines for the statistically efficient estimation of parameters of a modified Arrhenius model for chemical kinetics. A modified Arrhenius model is used for instance by modeling a flux of methane in troposphere or by chemical kinetics for reactions at membranes. D-optimal and filling designs for the Generalized Exponential Model with correlated observations are studied, considering the exponential covariance with or without nugget effect. Both equidistant and exact designs for small samples are examined, studying the behavior of different types of filling designs when a greater number of observations is preferred. Probably the main lesson we can learn is that the D-optimal design is analytically peculiar and these designs can be practically obtained only by numerical computation; however, specially two point locally D-optimal designs are very interesting, since they may help us to find a reasonable range for filling designs. The latter ones are probably only applicable when seeking for a higher number of design points. It is an interesting issue that very often the best designs do not use the whole design interval, but only a part of it; this should be taken into account by practitioners when they design their experiments. The second important observation is the large bias of the ML estimator of the correlation parameter. From the theoretical point of view this is not surprising since variance and correlation parameters are not simultaneously identifiable. We develop a bias reduction method and illustrate its effectiveness. We also provide practical implications for chemometrics.
Kernel k-nearest neighbor algorithm as a flexible SAR modeling tool
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Dong-Sheng Cao, Jian-Hua Huang, Jun Yan, Liang-Xiao Zhang, Qian-Nan Hu, Qing-Song Xu, Yi-Zeng Liang
A kernel version of k-nearest neighbor algorithm (k-NN) has been developed to model the complex relationship between molecular descriptors and bioactivities of compounds. Kernel k-NN is to perform the original k-NN algorithm by mapping the training samples in the input space into a high-dimensional feature space. It can be easily constructed by calculating the distance between samples in the feature space, directly deriving from the simple calculation of the kernel used. The developed kernel k-NN is very flexible to deal with complex nonlinear relationship, more importantly; it can also conveniently cope with some non-vectorial data only by the definition of different kernels. The results obtained from several real SAR datasets indicated that the performance of kernel k-NN is comparable to support vector machine methods. It can be regarded as an alternative modeling technique for several chemical problems including the study of structure–activity relationship (SAR). The source codes implementing kernel k-NN in R language are freely available at http://code.google.com/p/kernelmethods/.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Dong-Sheng Cao, Jian-Hua Huang, Jun Yan, Liang-Xiao Zhang, Qian-Nan Hu, Qing-Song Xu, Yi-Zeng Liang
A kernel version of k-nearest neighbor algorithm (k-NN) has been developed to model the complex relationship between molecular descriptors and bioactivities of compounds. Kernel k-NN is to perform the original k-NN algorithm by mapping the training samples in the input space into a high-dimensional feature space. It can be easily constructed by calculating the distance between samples in the feature space, directly deriving from the simple calculation of the kernel used. The developed kernel k-NN is very flexible to deal with complex nonlinear relationship, more importantly; it can also conveniently cope with some non-vectorial data only by the definition of different kernels. The results obtained from several real SAR datasets indicated that the performance of kernel k-NN is comparable to support vector machine methods. It can be regarded as an alternative modeling technique for several chemical problems including the study of structure–activity relationship (SAR). The source codes implementing kernel k-NN in R language are freely available at http://code.google.com/p/kernelmethods/.
Is sensing spatially distributed chemical information using sensory substitution with hyperspectral imaging possible?
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Bjørn K. Alsberg
Chemical images which are often computed from hyperspectral images contain the spatial distribution of chemical information of a scene. For many applications visualizing such images on computer screens is sufficient, however there are cases where there is a need to combine the chemical images more naturally with human vision. This is especially true for interactive work where chemical images are being rapidly updated to the user. Effective integration of spatial information in general from external sources with vision is a challenge. One approach is to overlay the view of the real physical world with computer-generated graphics as in augmented reality. However such cluttering of the visual field with computer-generated graphics may confuse the user and reduce functionality. Another is projecting the chemical images back onto the scene under study in order to render the chemical information in situ to the user. This approach, however has challenges in connection with very small and very large scenes under investigation. An alternative approach is here investigated based on the possibility of enhancing the human vision system using a sensory substitution device. Such devices enables a person to sense spatial information conveyed through sensory channels other than the eye, such as hearing and sense of touch. Results presented support the claim that spatial chemical information from a hyperspectral camera can be conveyed to the brain through a sensory channel different from the eyes. As this is tested on a sighted subject it effectively provides an extension of the human vision system to incorporate chemical information which otherwise is invisible to the naked eye.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Bjørn K. Alsberg
Chemical images which are often computed from hyperspectral images contain the spatial distribution of chemical information of a scene. For many applications visualizing such images on computer screens is sufficient, however there are cases where there is a need to combine the chemical images more naturally with human vision. This is especially true for interactive work where chemical images are being rapidly updated to the user. Effective integration of spatial information in general from external sources with vision is a challenge. One approach is to overlay the view of the real physical world with computer-generated graphics as in augmented reality. However such cluttering of the visual field with computer-generated graphics may confuse the user and reduce functionality. Another is projecting the chemical images back onto the scene under study in order to render the chemical information in situ to the user. This approach, however has challenges in connection with very small and very large scenes under investigation. An alternative approach is here investigated based on the possibility of enhancing the human vision system using a sensory substitution device. Such devices enables a person to sense spatial information conveyed through sensory channels other than the eye, such as hearing and sense of touch. Results presented support the claim that spatial chemical information from a hyperspectral camera can be conveyed to the brain through a sensory channel different from the eyes. As this is tested on a sighted subject it effectively provides an extension of the human vision system to incorporate chemical information which otherwise is invisible to the naked eye.
Handheld NIRS analysis for routine meat quality control: Database transfer from at-line instruments
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
E. Zamora-Rojas, D. Pérez-Marín, E. De Pedro-Sanz, J.E. Guerrero-Ginel, A. Garrido-Varo
Innovative advances in Near Infrared Spectroscopy (NIRS) instrumentation have enabled the development of new miniaturized spectrometers that combine NIRS technology with micro-electro-mechanical platforms, thus opening up new horizons for industrial NIRS applications. Many agro-food industries, laboratories and research centres already have large databases/libraries, built up over many years using NIR spectrometers; it is clearly important to preserve these data sets in order to avoid having to research and develop NIRS applications from scratch every time a new instrument appears on the market. Three standardization algorithms—Direct Standardization (DS), Piecewise Direct Standardization (PDS) and Spectral Difference by Wavelengths (SDW)—and varying numbers of standardization samples were evaluated for transferring meat quality databases from a high-performance at-line NIRS monochromator to a handheld based on micro-electro-mechanical systems (MEMS) NIRS spectrometer. The SDW algorithm and the use of 8 standardization samples yielded the best Standard Error of Prediction (SEP) values for the three chemical parameters transferred (0.72% for fat, 0.73% for moisture and 0.66% for protein). The successful transfer of the database to the MEMS-NIRS device enables a new approach for fast, low-cost, on-line/in-situ analysis of meat products.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
E. Zamora-Rojas, D. Pérez-Marín, E. De Pedro-Sanz, J.E. Guerrero-Ginel, A. Garrido-Varo
Innovative advances in Near Infrared Spectroscopy (NIRS) instrumentation have enabled the development of new miniaturized spectrometers that combine NIRS technology with micro-electro-mechanical platforms, thus opening up new horizons for industrial NIRS applications. Many agro-food industries, laboratories and research centres already have large databases/libraries, built up over many years using NIR spectrometers; it is clearly important to preserve these data sets in order to avoid having to research and develop NIRS applications from scratch every time a new instrument appears on the market. Three standardization algorithms—Direct Standardization (DS), Piecewise Direct Standardization (PDS) and Spectral Difference by Wavelengths (SDW)—and varying numbers of standardization samples were evaluated for transferring meat quality databases from a high-performance at-line NIRS monochromator to a handheld based on micro-electro-mechanical systems (MEMS) NIRS spectrometer. The SDW algorithm and the use of 8 standardization samples yielded the best Standard Error of Prediction (SEP) values for the three chemical parameters transferred (0.72% for fat, 0.73% for moisture and 0.66% for protein). The successful transfer of the database to the MEMS-NIRS device enables a new approach for fast, low-cost, on-line/in-situ analysis of meat products.
Extracting homologous series from mass spectrometry data by projection on predefined vectors
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Johan E. Carlson, James R. Gasson, Tanja Barth, Ingvar Eide
Multivariate statistical methods, such as Principal Component Analysis (PCA), have been used extensively over the past decades as tools for extracting significant information from complex data sets. As such they are very powerful and in combination with an understanding of underlying chemical principles, they have enabled researchers to develop useful models. A drawback with the methods is that they do not have the ability to incorporate any physical / chemical model of the system being studied during the statistical analysis. In this paper we present a method that can be used as a complement to traditional chemometric tools in finding patterns in mass spectrometry data. The method uses a pre-defined set of equally spaced sequences that are assumed to be present in the data. Allowing for some uncertainty in the peak locations due to the uncertainties for the measurement instrumentation, the measured spectra are then projected onto this set. It is shown that the resulting scores can be used to identify homologous series in measured mass spectra that differ significantly between different measured samples. As opposed to PCA, the loading vectors, in this case the pre-defined homologous series, are readily interpretable.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Johan E. Carlson, James R. Gasson, Tanja Barth, Ingvar Eide
Multivariate statistical methods, such as Principal Component Analysis (PCA), have been used extensively over the past decades as tools for extracting significant information from complex data sets. As such they are very powerful and in combination with an understanding of underlying chemical principles, they have enabled researchers to develop useful models. A drawback with the methods is that they do not have the ability to incorporate any physical / chemical model of the system being studied during the statistical analysis. In this paper we present a method that can be used as a complement to traditional chemometric tools in finding patterns in mass spectrometry data. The method uses a pre-defined set of equally spaced sequences that are assumed to be present in the data. Allowing for some uncertainty in the peak locations due to the uncertainties for the measurement instrumentation, the measured spectra are then projected onto this set. It is shown that the resulting scores can be used to identify homologous series in measured mass spectra that differ significantly between different measured samples. As opposed to PCA, the loading vectors, in this case the pre-defined homologous series, are readily interpretable.
Application of latent projective graph in variable selection for near infrared spectral analysis
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Xueguang Shao, Guorong Du, Ming Jing, Wensheng Cai
Latent projective graph (LPG) is a technique developed in chemical factor analysis (CFA) for investigating the nature of hyphenated data. Selective variables can be found because collinear variables present a straight line in an LPG. Variable selection in near infrared (NIR) spectral analysis has been a notoriously difficult task for improving the quality of the models, the aim of which is to find informative variables specific to the target component. In this work, based on the assumption that collinear wavelengths in the calibration spectra may have the same contribution to the modeling, LPG was adopted for variable selection in NIR spectral analysis. The variables located at the inflections of an LPG are found to be informative for the quantitative models. With three NIR datasets of pharmaceutical tablets, blood and plant samples, it was proved that a very parsimonious model can be built by using only several selected variables. Compared with the previous work, the method provides a simple way for variable selection.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Xueguang Shao, Guorong Du, Ming Jing, Wensheng Cai
Latent projective graph (LPG) is a technique developed in chemical factor analysis (CFA) for investigating the nature of hyphenated data. Selective variables can be found because collinear variables present a straight line in an LPG. Variable selection in near infrared (NIR) spectral analysis has been a notoriously difficult task for improving the quality of the models, the aim of which is to find informative variables specific to the target component. In this work, based on the assumption that collinear wavelengths in the calibration spectra may have the same contribution to the modeling, LPG was adopted for variable selection in NIR spectral analysis. The variables located at the inflections of an LPG are found to be informative for the quantitative models. With three NIR datasets of pharmaceutical tablets, blood and plant samples, it was proved that a very parsimonious model can be built by using only several selected variables. Compared with the previous work, the method provides a simple way for variable selection.
A QSAR study on the biodegradation activity of PAHs in aged contaminated sediments
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Xiang Xu, Xian-Guo Li, Shu-Wen Sun
The relationship between the chemical structure and biodegradation activity (−logt1/2) of 17 polycyclic aromatic hydrocarbons (PAHs) was studied using density functional theory (DFT) and stepwise multiple linear regression analysis (SMLR) methods. The equilibrium geometries and vibration frequency have been investigated at B3LYP/6-31+G(d,p) level. One high correlation coefficient was found between the wagging vibration frequency (Freq) of the whole molecule and −logt1/2, which is resulted by the special structural characteristic with a big conjugated system. By means of regression analysis, the main factors influencing biodegradation activity were screened, and the equations of quantitative structure–activity relationship (QSAR) were established. The evaluation of the developed QSAR showed that the relationships are significant and the model had good predictive ability. The QSAR model showed that the biodegradation activity was closely related to molecular structure: the chemical bond strength of benzene ring played an important role in biodegradation process; In addition, low molecular weight PAHs are more degradable than the high molecular weight compounds.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Xiang Xu, Xian-Guo Li, Shu-Wen Sun
The relationship between the chemical structure and biodegradation activity (−logt1/2) of 17 polycyclic aromatic hydrocarbons (PAHs) was studied using density functional theory (DFT) and stepwise multiple linear regression analysis (SMLR) methods. The equilibrium geometries and vibration frequency have been investigated at B3LYP/6-31+G(d,p) level. One high correlation coefficient was found between the wagging vibration frequency (Freq) of the whole molecule and −logt1/2, which is resulted by the special structural characteristic with a big conjugated system. By means of regression analysis, the main factors influencing biodegradation activity were screened, and the equations of quantitative structure–activity relationship (QSAR) were established. The evaluation of the developed QSAR showed that the relationships are significant and the model had good predictive ability. The QSAR model showed that the biodegradation activity was closely related to molecular structure: the chemical bond strength of benzene ring played an important role in biodegradation process; In addition, low molecular weight PAHs are more degradable than the high molecular weight compounds.
Post-experimental denoising and background subtraction of surface plasmon resonance images for better quantification
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Jun Chen, Yi Chen, Jiying Xu, Yiming Zhang, Tao Liao
A method was proposed to improve the quality of noise- and/or uneven background-degraded images obtained by surface plasmon resonance imaging experiments. The noise was suppressed by adaptive median filter in combination with wavelet transform, while the uneven background was flattened by subtraction with a three-dimensionally fitted surface. These operations improved the signal-to-noise ratio from 23.83dB to 41.36dB for real images and widened the quantitative linear range of picture gray value vs. concentration for about one order of magnitude, with linear correlation coefficient increased from 0.9558 to 0.9982. The method can be performed repeatedly until a better result is obtained and is thus cost-effective, highly competitive to experimental strategies and other computational methods.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Jun Chen, Yi Chen, Jiying Xu, Yiming Zhang, Tao Liao
A method was proposed to improve the quality of noise- and/or uneven background-degraded images obtained by surface plasmon resonance imaging experiments. The noise was suppressed by adaptive median filter in combination with wavelet transform, while the uneven background was flattened by subtraction with a three-dimensionally fitted surface. These operations improved the signal-to-noise ratio from 23.83dB to 41.36dB for real images and widened the quantitative linear range of picture gray value vs. concentration for about one order of magnitude, with linear correlation coefficient increased from 0.9558 to 0.9982. The method can be performed repeatedly until a better result is obtained and is thus cost-effective, highly competitive to experimental strategies and other computational methods.
Evaluation of the adsorption and rate constants of a photocatalytic degradation by means of HS-MCR-ALS. Study of process variables using experimental design
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Cristina Fernández, Anna de Juan, M. Pilar Callao, M. Soledad Larrechi
The effect of the catalyst type, the catalyst concentration and the pH on the global rate of the photocatalytic degradation of C.I. Acid Yellow 9 was studied. Adsorption and rate constants related to the physical adsorption on the catalyst surface and the degradation of the dye upon illumination were calculated by applying hybrid hard- and soft-multivariate curve resolution alternating least squares (HS-MCR-ALS), including the Langmuir–Hinshelwood kinetic model as hard restriction, to the UV–visible spectra recorded during the photocatalytic degradation process. The influence of the variables on the degradation rate was assessed using the experimental results obtained from a full factorial 23 experimental design. Physical adsorption was more relevant when TiO2 was employed as a catalyst. The global rate of the photocatalytic degradation of the dye was found to be closely related to the catalyst type, its concentration level and the interaction between both factors. pH was only found to be relevant when TiO2 was used. The photodegradation of C.I. Acid Yellow 9 was optimal when using high concentrations of ZnO as catalyst.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Cristina Fernández, Anna de Juan, M. Pilar Callao, M. Soledad Larrechi
The effect of the catalyst type, the catalyst concentration and the pH on the global rate of the photocatalytic degradation of C.I. Acid Yellow 9 was studied. Adsorption and rate constants related to the physical adsorption on the catalyst surface and the degradation of the dye upon illumination were calculated by applying hybrid hard- and soft-multivariate curve resolution alternating least squares (HS-MCR-ALS), including the Langmuir–Hinshelwood kinetic model as hard restriction, to the UV–visible spectra recorded during the photocatalytic degradation process. The influence of the variables on the degradation rate was assessed using the experimental results obtained from a full factorial 23 experimental design. Physical adsorption was more relevant when TiO2 was employed as a catalyst. The global rate of the photocatalytic degradation of the dye was found to be closely related to the catalyst type, its concentration level and the interaction between both factors. pH was only found to be relevant when TiO2 was used. The photodegradation of C.I. Acid Yellow 9 was optimal when using high concentrations of ZnO as catalyst.
New cluster mapping tools for the graphical assessment of non-dominated solutions in multi-objective optimization
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
R. Cela, M.H. Bollaín
Two new graphical tools for the interpretation of Pareto fronts and the selection of non-dominated solutions produced in multi-objective optimization processes (MOOPs), are presented. The first is a version of the parallel coordinates plots (PCP), modified by combining the PCP with the dendrogram representing the cluster analysis of non-dominated solutions in the decision variable space or in the objective space. A correspondence plot that simplifies interpretation of the above plots has also been developed. The second graphical tool is a cluster map (PFCM), produced by combining the information provided by the dendrograms calculated in the decision and the objective spaces, to provide a two-dimensional plot in which the non-dominated solutions are organized according to both dendrograms; the plot is colored on the basis of any of the objectives or a combination of these objectives when convenient. Two derived graphic tools consisting of a combination of the decision variables and the objectives and the dendrograms produced in the decision and the objective spaces have also been developed. All of these graphical tools are demonstrated with several mathematical functions available in the MOOP-related literature and with a real-world optimization process consisting of the computer-assisted method development of high-performance liquid chromatography.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
R. Cela, M.H. Bollaín
Two new graphical tools for the interpretation of Pareto fronts and the selection of non-dominated solutions produced in multi-objective optimization processes (MOOPs), are presented. The first is a version of the parallel coordinates plots (PCP), modified by combining the PCP with the dendrogram representing the cluster analysis of non-dominated solutions in the decision variable space or in the objective space. A correspondence plot that simplifies interpretation of the above plots has also been developed. The second graphical tool is a cluster map (PFCM), produced by combining the information provided by the dendrograms calculated in the decision and the objective spaces, to provide a two-dimensional plot in which the non-dominated solutions are organized according to both dendrograms; the plot is colored on the basis of any of the objectives or a combination of these objectives when convenient. Two derived graphic tools consisting of a combination of the decision variables and the objectives and the dendrograms produced in the decision and the objective spaces have also been developed. All of these graphical tools are demonstrated with several mathematical functions available in the MOOP-related literature and with a real-world optimization process consisting of the computer-assisted method development of high-performance liquid chromatography.
Online estimation of reject gas flow rates in compact flotation units for produced water treatment: A feasibility study
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Benjamin Kaku Arvoh, Steinar Asdahl, Karsten Rabe, Maths Halstensen
The largest waste water stream from oil and gas production wells is referred to as produced water (PW). One way of treating PW to limits acceptable for discharge into sea is by the use of compact flotation units (CFU). Currently, CFUs' are operated manually due to lack of advanced monitoring solutions. The main areas of interest in the current operation of the flotation unit are to provide online measurements of both liquid and gas flow rates through the reject stream of the flotation unit. A full scale feasibility study on the application of acoustic measurements and partial least squares regression as a tool for online estimation of the reject gas flow rate was investigated. From the experiments conducted to determine the optimal sensor location, it was concluded that there were no significant differences between the four sensor locations investigated. Several reject gas flow rate models were calibrated and validated with fully independent data. The average root mean square error of prediction (RMSEP) was 7.5% within the experimental range (0.07–2.5 Sm3/h). The RMSEP for experiments with varying salt concentration was 0.25% (within the range of 0–5.5%) whilst that for varying temperature was 1.8°C (within the range of 30–60°C). The model for reject liquid flow rates through the CFU had a RMSEP of 16.86l/h (within the range of 225–485l/h). These promising results will form the basis for further development and implementation of the technique in flotation units.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Benjamin Kaku Arvoh, Steinar Asdahl, Karsten Rabe, Maths Halstensen
The largest waste water stream from oil and gas production wells is referred to as produced water (PW). One way of treating PW to limits acceptable for discharge into sea is by the use of compact flotation units (CFU). Currently, CFUs' are operated manually due to lack of advanced monitoring solutions. The main areas of interest in the current operation of the flotation unit are to provide online measurements of both liquid and gas flow rates through the reject stream of the flotation unit. A full scale feasibility study on the application of acoustic measurements and partial least squares regression as a tool for online estimation of the reject gas flow rate was investigated. From the experiments conducted to determine the optimal sensor location, it was concluded that there were no significant differences between the four sensor locations investigated. Several reject gas flow rate models were calibrated and validated with fully independent data. The average root mean square error of prediction (RMSEP) was 7.5% within the experimental range (0.07–2.5 Sm3/h). The RMSEP for experiments with varying salt concentration was 0.25% (within the range of 0–5.5%) whilst that for varying temperature was 1.8°C (within the range of 30–60°C). The model for reject liquid flow rates through the CFU had a RMSEP of 16.86l/h (within the range of 225–485l/h). These promising results will form the basis for further development and implementation of the technique in flotation units.
New similarity metrics for Raman spectroscopy
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Shehroz S. Khan, Michael G. Madden
Similarity metrics are at the heart of spectral library search procedures that are used to identify unknown substances. The problem is relatively easy when the query spectrum (that is, a spectrum of the substance to be identified) is directly represented in the library, but in general this is not the case, and the query spectrum may come from a mixture of substances that are either individually represented in the library or as a mixture. In such cases, employing standard search metrics may not yield good results. A well-known general strategy to improve search is to design domain-specific metrics that capture its intrinsic properties. In this paper, we present a new Raman spectroscopy specific spectral similarity metric, Spectral Linear Kernel, which captures the domain subtleties while performing spectral search and performs better in comparison to standard spectral search methods. We also present a new modified Euclidean measure which not only performs better than the standard Euclidean method but other standard methods. We evaluate our results on Raman spectroscopy data for chlorinated solvents.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Shehroz S. Khan, Michael G. Madden
Similarity metrics are at the heart of spectral library search procedures that are used to identify unknown substances. The problem is relatively easy when the query spectrum (that is, a spectrum of the substance to be identified) is directly represented in the library, but in general this is not the case, and the query spectrum may come from a mixture of substances that are either individually represented in the library or as a mixture. In such cases, employing standard search metrics may not yield good results. A well-known general strategy to improve search is to design domain-specific metrics that capture its intrinsic properties. In this paper, we present a new Raman spectroscopy specific spectral similarity metric, Spectral Linear Kernel, which captures the domain subtleties while performing spectral search and performs better in comparison to standard spectral search methods. We also present a new modified Euclidean measure which not only performs better than the standard Euclidean method but other standard methods. We evaluate our results on Raman spectroscopy data for chlorinated solvents.
Recognition of the hardness of licorice seeds using a semi-supervised learning method and near-infrared spectral data
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Liming Yang, Qun Sun
The recognition of the hardness of licorice seeds is a challenging task. The purpose of this investigation is to identify the hardness of licorice seeds employing a semi-supervised learning method and near-infrared spectroscopy. An excellent semi-supervised learning model, the semi-supervised support vector machine (S3VM), is built using the small labeled samples and the large unlabeled samples. Moreover, the proposed model is solved by employing an effective method, the robust DC (difference of convex functions) programming. The resulting algorithm only requires the solving of a few linear programs. Furthermore, this model is used for the direct classification of licorice samples. Comparing with the supervised support vector machine (SVM), experimental results on different spectral regions show that incorporating unlabeled samples in training improves the generalization when insufficient training information is available. Moreover, our method outperforms the existing S3VM method by obtaining better performance in different spectral regions. These results show that it is possible to identify the hardness of licorice seeds using the proposed S3VM and near-infrared spectroscopic data. We hope that the results obtained in this study will help further investigations of the hardness of crop seeds.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Liming Yang, Qun Sun
The recognition of the hardness of licorice seeds is a challenging task. The purpose of this investigation is to identify the hardness of licorice seeds employing a semi-supervised learning method and near-infrared spectroscopy. An excellent semi-supervised learning model, the semi-supervised support vector machine (S3VM), is built using the small labeled samples and the large unlabeled samples. Moreover, the proposed model is solved by employing an effective method, the robust DC (difference of convex functions) programming. The resulting algorithm only requires the solving of a few linear programs. Furthermore, this model is used for the direct classification of licorice samples. Comparing with the supervised support vector machine (SVM), experimental results on different spectral regions show that incorporating unlabeled samples in training improves the generalization when insufficient training information is available. Moreover, our method outperforms the existing S3VM method by obtaining better performance in different spectral regions. These results show that it is possible to identify the hardness of licorice seeds using the proposed S3VM and near-infrared spectroscopic data. We hope that the results obtained in this study will help further investigations of the hardness of crop seeds.
WSPLS — A new approach towards mixture modeling and accelerated product development
07 May 2012,
20:50:54
Publication year:
2012
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Salvador García-Muñoz, Mark Polizzi
A new method is presented to model mixture data which simultaneously regresses the fractions of the materials used in a series of blends, and the matrix of the physical properties of the materials used in such blends to the properties measured from the resulting blend. The Weighted Scores Projection to Latent Structures (WSPLS) method combines the fractions of the used materials and their physical properties by first transforming the physical properties with a Principal Component Analysis (PCA) model and then estimating a matrix of weighted average scores using the fractions of the materials used and the corresponding scores for each material from the PCA models. This matrix of weighted scores is the regressor in the PLS model against the measured properties of the mixture. The new method is contrasted with other alternatives and shown to provide robust models with strong predictive components across all latent variables. A data set from blends of pharmaceutical powders is used to illustrate the features of the method proposed.
Source:Chemometrics and Intelligent Laboratory Systems, Volume 114
Salvador García-Muñoz, Mark Polizzi
A new method is presented to model mixture data which simultaneously regresses the fractions of the materials used in a series of blends, and the matrix of the physical properties of the materials used in such blends to the properties measured from the resulting blend. The Weighted Scores Projection to Latent Structures (WSPLS) method combines the fractions of the used materials and their physical properties by first transforming the physical properties with a Principal Component Analysis (PCA) model and then estimating a matrix of weighted average scores using the fractions of the materials used and the corresponding scores for each material from the PCA models. This matrix of weighted scores is the regressor in the PLS model against the measured properties of the mixture. The new method is contrasted with other alternatives and shown to provide robust models with strong predictive components across all latent variables. A data set from blends of pharmaceutical powders is used to illustrate the features of the method proposed.
No comments:
Post a Comment