The Experts below are selected from a list of 8445 Experts worldwide ranked by ideXlab platform
J C Noyer - One of the best experts on this subject based on the ideXlab platform.
-
formulating robust linear regression estimation as a one class lda criterion discriminative Hat Matrix
2013Co-Authors: Franck Dufrenois, J C NoyerAbstract:Linear discriminant analysis, such as Fisher's criterion, is a statistical learning tool traditionally devoted to separating a training dataset into two or even several classes by the way of linear decision boundaries. In this paper, we show tHat this tool can formalize the robust linear regression problem as a robust estimator will do. More precisely, we develop a one-class Fischer's criterion in which the maximization provides both the regression parameters and the separation of the data in two classes: typical data and atypical data or outliers. This new criterion is built on the statistical properties of the subspace decomposition of the Hat Matrix. From this angle, we improve the discriminative properties of the Hat Matrix which is traditionally used as outlier diagnostic measure in linear regression. Naturally, we call this new approach discriminative Hat Matrix. The proposed algorithm is fully nonsupervised and needs only the initialization of one parameter. Synthetic and real datasets are used to study the performance both in terms of regression and classification of the proposed approach. We also illustrate its potential application to image recognition and fundamental Matrix estimation in computer vision.
-
discriminative kernel Hat Matrix a new tool for automatic outlier identification
2012Co-Authors: Franck Dufrenois, J C NoyerAbstract:Identifying outlying observations in data sets is one of the classical topics in robust statistics. We propose to solve this problem by a new one-class kernel Fisher criterion based on the statistics of the subspace decomposition of the kernel Hat Matrix diagonal. This work extends the recent study proposed in [1] to the nonlinear case. We show here tHat the maximization of this new contrast measure comes down to search an optimal projection subspace and an optimal indicator Matrix. Next, we derive a separating boundary between the dominant population and outliers. We show tHat the maximum of the criterion corresponds both to an optimal value of the kernel parameters and to an optimal classification providing the expected fraction of outliers. This means tHat these two problems are intimately related. Synthetic and real data sets are used to study the performance of the proposed approach.
-
discriminative Hat Matrix a new tool for outlier identification and linear regression
2011Co-Authors: Franck Dufrenois, J C NoyerAbstract:The Hat Matrix is an important auxiliary quantity in linear regression theory for detecting errors in predictors. Traditionally, the comparison of the diagonal elements with a calibration point serves as decision rule for separating a dominant linear population from outliers. However, several problems exist: first, the calibration point is not well defined because no exact statistical distribution (asymptotic form) of the Hat Matrix diagonal exists [1]. Secondly, being based on the standard covariance Matrix, this outlying measure looses its efficiency when the rate of “atypical” observations becomes large [2][3]. In this paper, we present a discriminative version of the Hat Matrix (DHM) which transposes this classification problem into a subspace clustering problem. We propose a linear discriminant analysis based criterion directly built on the properties of the Hat Matrix and we show tHat its maximization leads to search an optimal projection subspace and an optimal indicator Matrix. We also show tHat the statistic of the Hat Matrix diagonal “projected” on this optimal subspace has an exact χ2 behaviour and thus makes it possible to identify outliers by way of hyptothesis testing. Synthetic data sets are used to study the performance both in terms of regression and classification of the proposed approach. We also illustrate its potential application to motion segmentation in image sequences.
-
a kernel Hat Matrix based rejection criterion for outlier removal in support vector regression
2009Co-Authors: Franck Dufrenois, J C NoyerAbstract:In this paper, we propose a kernel Hat Matrix based learning stage for outlier removal. In particular, we show tHat the gaussian kernel Hat Matrix have very interesting discriminative properties under the condition of choosing appropriate values for kernel parameters. Thus, we develop a practical model selection criteria in order to well separate the “outlier” distribution from the “dominant” distribution. This learning stage, beforehand applied to the training data set, offers a new answer for down-weighting outliers corrupting both the response and predictor variables in regression tasks. The application to simulated and real data shows the robustness of the proposed approach.
Franck Dufrenois - One of the best experts on this subject based on the ideXlab platform.
-
formulating robust linear regression estimation as a one class lda criterion discriminative Hat Matrix
2013Co-Authors: Franck Dufrenois, J C NoyerAbstract:Linear discriminant analysis, such as Fisher's criterion, is a statistical learning tool traditionally devoted to separating a training dataset into two or even several classes by the way of linear decision boundaries. In this paper, we show tHat this tool can formalize the robust linear regression problem as a robust estimator will do. More precisely, we develop a one-class Fischer's criterion in which the maximization provides both the regression parameters and the separation of the data in two classes: typical data and atypical data or outliers. This new criterion is built on the statistical properties of the subspace decomposition of the Hat Matrix. From this angle, we improve the discriminative properties of the Hat Matrix which is traditionally used as outlier diagnostic measure in linear regression. Naturally, we call this new approach discriminative Hat Matrix. The proposed algorithm is fully nonsupervised and needs only the initialization of one parameter. Synthetic and real datasets are used to study the performance both in terms of regression and classification of the proposed approach. We also illustrate its potential application to image recognition and fundamental Matrix estimation in computer vision.
-
discriminative kernel Hat Matrix a new tool for automatic outlier identification
2012Co-Authors: Franck Dufrenois, J C NoyerAbstract:Identifying outlying observations in data sets is one of the classical topics in robust statistics. We propose to solve this problem by a new one-class kernel Fisher criterion based on the statistics of the subspace decomposition of the kernel Hat Matrix diagonal. This work extends the recent study proposed in [1] to the nonlinear case. We show here tHat the maximization of this new contrast measure comes down to search an optimal projection subspace and an optimal indicator Matrix. Next, we derive a separating boundary between the dominant population and outliers. We show tHat the maximum of the criterion corresponds both to an optimal value of the kernel parameters and to an optimal classification providing the expected fraction of outliers. This means tHat these two problems are intimately related. Synthetic and real data sets are used to study the performance of the proposed approach.
-
discriminative Hat Matrix a new tool for outlier identification and linear regression
2011Co-Authors: Franck Dufrenois, J C NoyerAbstract:The Hat Matrix is an important auxiliary quantity in linear regression theory for detecting errors in predictors. Traditionally, the comparison of the diagonal elements with a calibration point serves as decision rule for separating a dominant linear population from outliers. However, several problems exist: first, the calibration point is not well defined because no exact statistical distribution (asymptotic form) of the Hat Matrix diagonal exists [1]. Secondly, being based on the standard covariance Matrix, this outlying measure looses its efficiency when the rate of “atypical” observations becomes large [2][3]. In this paper, we present a discriminative version of the Hat Matrix (DHM) which transposes this classification problem into a subspace clustering problem. We propose a linear discriminant analysis based criterion directly built on the properties of the Hat Matrix and we show tHat its maximization leads to search an optimal projection subspace and an optimal indicator Matrix. We also show tHat the statistic of the Hat Matrix diagonal “projected” on this optimal subspace has an exact χ2 behaviour and thus makes it possible to identify outliers by way of hyptothesis testing. Synthetic data sets are used to study the performance both in terms of regression and classification of the proposed approach. We also illustrate its potential application to motion segmentation in image sequences.
-
a kernel Hat Matrix based rejection criterion for outlier removal in support vector regression
2009Co-Authors: Franck Dufrenois, J C NoyerAbstract:In this paper, we propose a kernel Hat Matrix based learning stage for outlier removal. In particular, we show tHat the gaussian kernel Hat Matrix have very interesting discriminative properties under the condition of choosing appropriate values for kernel parameters. Thus, we develop a practical model selection criteria in order to well separate the “outlier” distribution from the “dominant” distribution. This learning stage, beforehand applied to the training data set, offers a new answer for down-weighting outliers corrupting both the response and predictor variables in regression tasks. The application to simulated and real data shows the robustness of the proposed approach.
Amir H Mohammadi - One of the best experts on this subject based on the ideXlab platform.
-
a reliable model for estimating the wax deposition rate during crude oil production and processing
2014Co-Authors: Arash Kamari, Amir H Mohammadi, Alireza Bahadori, Sohrab ZendehboudiAbstract:Deposition of wax in surface and subsurface pipes and even perforations can lead to serious problems including pore spaces blockage, plugging of pipelines, and minimum profitability. Therefore, understanding wax deposition and wax related-properties improves the oil recovery and processing operations in petroleum industry. In this research work, the rate of wax deposition is correlated to a number of main parameters such as the dynamic viscosity of crude oil, shear stress, gradient of wax molecular concentration and temperature difference in pipeline system through implementation of a newly developed model, known as least squares support vector machine (LSSVM) along with the coupled simulated annealing (CSA) optimization strategy. The possible outliers are detected through employing the Leverage technique which involves residual errors plots, Williams' plot, and Hat Matrix. The results imply tHat whole collected real data are in applicability domain of the proposed model. Using a comprehensive statistical...
-
asphaltene precipitation due to natural depletion of reservoir determination using a sara fraction based intelligent model
2013Co-Authors: Farhad Gharagheizi, Abdolhossein Hemmatisarapardeh, Reza Alipouryeganehmarand, Ali Naseri, Anoush Safiabadi, Poorandokht Ilanikashkouli, Amir H MohammadiAbstract:Abstract Precipitation of asphaltene leads to rigorous problems in petroleum industry such as: wettability alterations, relative permeability reduction, blockage of the flow with additional pressure drop in wellbore tubing, upstream process facilities and surface pipelines. Experimentally determination of the asphaltene precipitation is costly and time consuming. Therefore, searching for some other quick and accurate methods for determination of the asphaltene precipitation is inevitable. The objective of this communication is to present a reliable and predictive model namely, the least – squares support vector machine (LSSVM) to predict the asphaltene precipitation. This model has been developed and tested using 157 series of experimental data for 32 different crude oils from a number of Iranian oil reservoirs. The ranges of data used to develop the model cover many of Iranian oil reservoirs PVT data and consequently the developed model could be reliable for prediction of other Iranian oil reservoirs’ samples. Statistical and graphical error analysis have been carried out to establish the adequacy and accuracy of the model. The results show tHat the developed model provides predictions in good agreement with experimental data. Furthermore, it is illustrated tHat the proposed method is capable of simulating the actual physical trend of the asphaltene precipitation with variation of pressure. Finally, the Leverage approach, in which the statistical Hat Matrix, Williams plot, and the residuals of the model results lead to identification of the likely outliers, has been performed. Fortunately, all the experimental data seem to be reliable except five. Thus, the developed model could be reliable for prediction of the asphaltene precipitation in its applicability domain. This model can be implemented in any reservoir simulator software and it provides enough accuracy and performance over the existing methods.
-
a novel method for evaluation of asphaltene precipitation titration data
2012Co-Authors: Amir H Mohammadi, Ali Eslamimanesh, Farhad Gharagheizi, Dominique RichonAbstract:In this work, we propose a mathematical method for detection of the probable doubtful asphaltene precipitation titration data. The algorithm is performed on the basis of the Leverage approach, in which the statistical Hat Matrix, Williams Plot, and the residuals of the model results lead to identify the probable outliers. This method not only contributes to outliers diagnostics but also defines the range of applicability of the applied models and quality of the existing experimental data. Two available scaling equations from the literature are used to pursue the calculation steps. It is found from the obtained results tHat: I. The applied models to represent/predict the weight percent of asphaltene precipitation are statistically valid and correct. II. All the treated experimental titration data seem to be reliable except one. III. The whole data points present in the dataset are within the domain of applicability of the employed models.
Dominique Richon - One of the best experts on this subject based on the ideXlab platform.
-
a novel method for evaluation of asphaltene precipitation titration data
2012Co-Authors: Amir H Mohammadi, Ali Eslamimanesh, Farhad Gharagheizi, Dominique RichonAbstract:In this work, we propose a mathematical method for detection of the probable doubtful asphaltene precipitation titration data. The algorithm is performed on the basis of the Leverage approach, in which the statistical Hat Matrix, Williams Plot, and the residuals of the model results lead to identify the probable outliers. This method not only contributes to outliers diagnostics but also defines the range of applicability of the applied models and quality of the existing experimental data. Two available scaling equations from the literature are used to pursue the calculation steps. It is found from the obtained results tHat: I. The applied models to represent/predict the weight percent of asphaltene precipitation are statistically valid and correct. II. All the treated experimental titration data seem to be reliable except one. III. The whole data points present in the dataset are within the domain of applicability of the employed models.
Farhad Gharagheizi - One of the best experts on this subject based on the ideXlab platform.
-
asphaltene precipitation due to natural depletion of reservoir determination using a sara fraction based intelligent model
2013Co-Authors: Farhad Gharagheizi, Abdolhossein Hemmatisarapardeh, Reza Alipouryeganehmarand, Ali Naseri, Anoush Safiabadi, Poorandokht Ilanikashkouli, Amir H MohammadiAbstract:Abstract Precipitation of asphaltene leads to rigorous problems in petroleum industry such as: wettability alterations, relative permeability reduction, blockage of the flow with additional pressure drop in wellbore tubing, upstream process facilities and surface pipelines. Experimentally determination of the asphaltene precipitation is costly and time consuming. Therefore, searching for some other quick and accurate methods for determination of the asphaltene precipitation is inevitable. The objective of this communication is to present a reliable and predictive model namely, the least – squares support vector machine (LSSVM) to predict the asphaltene precipitation. This model has been developed and tested using 157 series of experimental data for 32 different crude oils from a number of Iranian oil reservoirs. The ranges of data used to develop the model cover many of Iranian oil reservoirs PVT data and consequently the developed model could be reliable for prediction of other Iranian oil reservoirs’ samples. Statistical and graphical error analysis have been carried out to establish the adequacy and accuracy of the model. The results show tHat the developed model provides predictions in good agreement with experimental data. Furthermore, it is illustrated tHat the proposed method is capable of simulating the actual physical trend of the asphaltene precipitation with variation of pressure. Finally, the Leverage approach, in which the statistical Hat Matrix, Williams plot, and the residuals of the model results lead to identification of the likely outliers, has been performed. Fortunately, all the experimental data seem to be reliable except five. Thus, the developed model could be reliable for prediction of the asphaltene precipitation in its applicability domain. This model can be implemented in any reservoir simulator software and it provides enough accuracy and performance over the existing methods.
-
a novel method for evaluation of asphaltene precipitation titration data
2012Co-Authors: Amir H Mohammadi, Ali Eslamimanesh, Farhad Gharagheizi, Dominique RichonAbstract:In this work, we propose a mathematical method for detection of the probable doubtful asphaltene precipitation titration data. The algorithm is performed on the basis of the Leverage approach, in which the statistical Hat Matrix, Williams Plot, and the residuals of the model results lead to identify the probable outliers. This method not only contributes to outliers diagnostics but also defines the range of applicability of the applied models and quality of the existing experimental data. Two available scaling equations from the literature are used to pursue the calculation steps. It is found from the obtained results tHat: I. The applied models to represent/predict the weight percent of asphaltene precipitation are statistically valid and correct. II. All the treated experimental titration data seem to be reliable except one. III. The whole data points present in the dataset are within the domain of applicability of the employed models.
