DUPLEX Algorithm Research Articles

Optimal Sample Size for Predicting Viability of Cabbage and Radish Seeds Based on near Infrared Spectra of Single Seeds

The effects of the number of seeds in a training sample set on the ability to predict the viability of cabbage or radish seeds are presented and discussed. The supervised classification method extended canonical variates analysis (ECVA) was used to develop a classification model. Calibration sub-sets of different sizes were chosen randomly with several iterations and using the spectral-based sample selection algorithms DUPLEX and CADEX. An independent test set was used to validate the developed classification models. The results showed that 200 seeds were optimal in a calibration set for both cabbage and radish data. The misclassification rates at optimal sample size were 8%, 6% and 7% for cabbage and 3%, 3% and 2% for radish respectively for random method (averaged for 10 iterations), DUPLEX and CADEX algorithms. This was similar to the misclassification rate of 6% and 2% for cabbage and radish obtained using all 600 seeds in the calibration set. Thus, the number of seeds in the calibration set can be reduced by up to 67% without significant loss of classification accuracy, which will effectively enhance the cost-effectiveness of NIR spectral analysis. Wavelength regions important for the discrimination between viable and non-viable seeds were identified using interval ECVA (iECVA) models, ECVA weight plots and the mean difference spectrum for viable and non-viable seeds.

On the selection of samples for multivariate regression analysis: application to near-infrared (NIR) calibration models for the prediction of pulp yield in Eucalyptus nitens

The effects of using reduced calibration sets on the development of near-infrared (NIR) calibration models for the prediction of kraft pulp yield in Eucalyptus nitens (Dean & Maiden) Maiden trees were explored. Three selection techniques based on NIR spectral data (CADEX (computer-aided design of experiments), DUPLEX, and SELECT algorithms) and one selection method based on a measured property (RANKING algorithm) were used for analysis and compared against a model using all data. The effect of using calibration sets of different sizes was also evaluated. All sample-selection methods resulted in models of similar performance compared with the model fitted using all samples. For calibration purposes, RANKING selection resulted in models with the lowest errors of cross-validation, followed by the DUPLEX, CADEX, and SELECT methods. In terms of validation, the SELECT and CADEX methods resulted in lower errors of prediction compared with the DUPLEX and RANKING algorithms. In general, cross-validation and prediction errors decreased as the number of calibration samples increased. These results show that it is possible to obtain adequate NIR calibration models with a reduced number of samples allowing the remaining samples to be used for model validation and that sample selection based on NIR spectral data alone is as successful as selection based on a measured property.

Discrimination of biofilm samples using pattern recognition techniques

Biofilms are complex aggregates formed by microorganisms such as bacteria, fungi and algae, which grow at the interfaces between water and natural or artificial materials. They are actively involved in processes of sorption and desorption of metal ions in water and reflect the environmental conditions in the recent past. Therefore, biofilms can be used as bioindicators of water quality. The goal of this study was to determine whether the biofilms, developed in different aquatic systems, could be successfully discriminated using data on their elemental compositions. Biofilms were grown on natural or polycarbonate materials in flowing water, standing water and seawater bodies. Using an unsupervised technique such as principal component analysis (PCA) and several supervised methods like classification and regression trees (CART), discriminant partial least squares regression (DPLS) and uninformative variable elimination–DPLS (UVE-DPLS), we could confirm the uniqueness of sea biofilms and make a distinction between flowing water and standing water biofilms. The CART, DPLS and UVE-DPLS discriminant models were validated with an independent test set selected either by the Kennard and Stone method or the duplex algorithm. The best model was obtained from CART with 100% correct classification rate for the test set designed by the Kennard and Stone algorithm. With CART, one variable describing the Mg content in the biofilm water phase was found to be important for the discrimination of flowing water and standing water biofilms.

Prediction of ozone tropospheric degradation rate constants by projection pursuit regression

Quantitative structure–property relationship (QSPR) models were developed to predict degradation rate constants of ozone tropospheric and to study the degradation reactivity mechanism of 116 diverse compounds. DUPLEX algorithm was utilized to design the training and test sets. Seven molecular descriptors selected by the heuristic method (HM) were used as inputs to perform multiple linear regression (MLR), support vector machine (SVM) and projection pursuit regression (PPR) studies. The PPR model performs best both in the fitness and in the prediction capacity. For the test set, it gave a predictive correlation coefficient ( R) of 0.955, root mean square error (RMSE) of 1.041 and absolute average relative deviation (AARD, %) of 4.663, respectively. The results proved that PPR is a useful tool that can be used to solve the nonlinear problems in QSPR. In addition, methods used in this paper are simple, practical and effective for chemists to predict the ozone degradation rate constants of compounds in troposphere.

Genetic algorithm for controllers in elevator groups: analysis and simulation during lunchpeak traffic

Abstract A genetic algorithm (GAHCA) is proposed to control elevator groups of professional buildings. The genetic algorithm is compared with the universal controller algorithm in industry applications. In order to do so an ARENA simulation scenario has been generated during heavy lunchpeak traffic conditions. The results allow us to affirm that our genetic algorithm reaches a better performance attending to the system waiting times than traditional duplex algorithms.

Russian paint output drops by more than 10%

This paper describes the construction of a QSAR model to relate the structures of various derivatives of neocryptolepine to their anti-malarial activities. QSAR classification models were build using Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Classification and Regression Trees (CART), Partial Least Squares – Discriminant Analysis (PLS-DA), Orthogonal Projections to Latent Structures – Discriminant Analysis (OPLS-DA), and Support Vector Machines for Classification (SVM-C), using four sets of molecular descriptors as explanatory variables. Prior to classification, the molecules were divided into a training and a test set using the duplex algorithm. The different classification models were compared regarding their predictive ability, simplicity, and interpretability. Both binary and multi-class classification models were constructed. For classification into three classes, CART and One-Against-One (OAO)-SVM-C were found to be the best predictive methods, while for classification into two classes, LDA, QDA and CART were.

Cross-Validation and Information Criteria in Causal Modeling

Many applications of causal modeling in marketing involve selection among several competing causal models. The author investigates whether common criteria for model selection such as cross-validation indices and information criteria are likely to lead to discovery of the correct population model. Guidance on the use of these selection criteria in practice is provided for substantive marketing researchers. Results indicate that the adequacy of cross-validation depends critically on the method used for sample-splitting. The author suggests the application of Snee's DUPLEX algorithm in this context. For situations in which the assumption of multinormally distributed variables is justified, information criteria are found to be highly appropriate for model selection, outperforming cross-validation methods in several respects.

Validation of Regression Models: Methods and Examples

Methods to determine the validity of regression models include comparison of model predictions and coefficients with theory, collection of new data to check model predictions. comparison of results with theoretical model calculations, and data splitting or cross-validation in which a portion of the data is used to estimate the model coefficients, and the remainder of the data is used to measure the prediction accuracy of the model. An expository review of these methods is presented. It is concluded that data splitting is an effective method of model validation when it is not practical to collect new data to test the model. The DUPLEX algorithm, developed by R. W. Kennard, is recommended for dividing the data into the estimation set and prediction set when there is no obvious variable such as time to use as a basis to split the data. Several examples are included to illustrate the various methods of model validation.

Validation of Regression Models: Methods and Examples

Methods to determine the validity of regression models include comparison of model predictions and coefficients with theory, collection of new data to check model predictions. comparison of results with theoretical model calculations, and data splitting or cross-validation in which a portion of the data is used to estimate the model coefficients, and the remainder of the data is used to measure the prediction accuracy of the model. An expository review of these methods is presented. It is concluded that data splitting is an effective method of model validation when it is not practical to collect new data to test the model. The DUPLEX algorithm, developed by R. W. Kennard, is recommended for dividing the data into the estimation set and prediction set when there is no obvious variable such as time to use as a basis to split the data. Several examples are included to illustrate the various methods of model validation.