Combination Of Principal Component Analysis Research Articles

Polycyclic aromatic hydrocarbons and organochlorine pesticides in rice hull from a typical e-waste recycling area in southeast China: temporal trend, source, and exposure assessment

The residue levels of 16 US EPA priority polycyclic aromatic hydrocarbons (PAHs) and 16 selected organochlorine pesticides (OCPs) in rice and rice hull collected from a typical e-waste recycling area in southeast China were investigated from 2005 to 2007. PAHs and OCPs also were measured in ten mollusk species (soft tissues) collected in an adjacent bay in 2007. Individual PAHs were frequently found in the entire sample set (including the rice, hull, and mollusk samples) with a detection rate of 73%. The total concentrations of 16 PAHs (ΣPAHs) and 16 OCPs (ΣOCPs) were in the range of 40.8-432ng/g dry weight (mean: 171ng/g) and 2.35-925ng/g (122ng/g), respectively, which were comparable or higher than those reported in some polluted areas. Statistical comparisons suggested that the concentrations of contaminants in hull gradually decreased from 2005 to 2007 and the residue levels were generally in the order of mollusk, hull, and rice, on a dry weight basis. Principal component analysis in combination with diagnostic ratios implied that combustion of coal, wood, and plastic wastes that are closely associated with crude e-waste recycling activities is the main source of PAHs. The finding of decreasing trend of concentrations of PAHs in this area is consistent with the efforts of local authorities to strengthen regulations on illegal e-waste recycling activities. Composition analysis suggested that there is a recent usage or discharge of hexachlorocyclohexane and dichlorodiphenyltrichloroethane into the tested area. The estimated daily intake (EDI) of ΣPAHs and ΣOCPs (calculated from mean concentrations) through rice and mollusk consumption was 0.411 and 0.921μg/kg body weight (bw)/day, respectively.

Forecasting retained earnings of privately held companies with PCA and L1 regression

We use proprietary data collected by SVB Analytics, an affiliate of Silicon Valley Bank, to forecast the retained earnings of privately held companies. Combining methods of principal component analysis (PCA) and L1/quantile regression, we build multivariate linear models that feature excellent in‐sample fit and strong out‐of‐sample predictive accuracy. The combined PCA and L1 technique effectively deals with multicollinearity and non‐normality of the data, and also performs favorably when compared against a variety of other models. Additionally, we propose a variable ranking procedure that explains which variables from the current quarter are most predictive of the next quarter's retained earnings. We fit models to the top five variables identified by the ranking procedure and thereby, discover interpretable models with excellent out‐of‐sample performance. Copyright © 2013 John Wiley & Sons, Ltd.

Implementation and evaluation of multivariate analysis for groundwater hydrochemistry assessment in arid environments: a case study of Hajeb Elyoun–Jelma, Central Tunisia

In most arid zones, groundwater (GW) is the major source of domestic, agriculture, drinking, and industrial water. Accordingly, the monitoring of its quality by different techniques and tools is a vital issue. The purpose of this paper is the evaluation of the combination of principal components analysis (PCA) and geostatistics as a technique for (1) identifying the processes affecting the groundwater chemistry of the detrital unconfined Middle Miocene Aquifer (MMA) of the Hajeb elyoun Jelma (HJ) aquifer (Tunisia) and (2) mapping the controlling variables for groundwater quality. This work is based on a limited database recorded in 22 wells tapping the aquifer and unequally distributed in the field. The proposed approach is carried out in two steps. In the first step, the application of PCA revealed that rock–water interaction, agriculture irrigation and domestic effluents could explain 85 % of the observed variability of the chemical GW quality of the MMA. As a result, two new variables are defined: V1 (rock–water interaction influence) and V2 (irrigation and domestic effluent influence). In the second step, the spatial variability of these variables over the extent of the MMA is mapped by applying a kriging interpolation technique. The results of this study suggest that, while both natural and anthropogenic processes contribute to the GW quality of the MMA, natural impacts can be considered as the most important.

Estimation of Petroleum Reservoir Parameters Using an Integrated Approach Neural Network, Principal Component Analysis and Fisher Discriminant Analysis

A new predictive methodology is introduced, based on a combined principal component analysis (PCA), Fisher discriminant analysis (FDA), and artificial neural network (ANN) methodologies for parameters estimation of a petroleum reservoir. Prediction of continuous petrophysical parameters is often time consuming and complicated because of geological variability such as facies changes due to sedimentary and structural changes. The petrophysical parameters, however, are usually difficult to measure due to reliability considerations, limitations insights on cost, inappropriate instrument maintenance, and sensor failures, evaluated by crude diagrams of reservoir parameters valuably. PCA and FDA provides an optimal lower dimensional representation in terms of discriminating among classes of data and are developed utilizing the reservoir historical data to incorporate reliability and prediction capabilities of ANN. The developed soft sensors are applied to predict the parameters of Marun reservoir located in Ahwaz, Iran, by utilizing the available geophysical well log data. The resulting outcomes demonstrate the promising capabilities of the proposed hybrid PCA-FDA-NN methodology than the conventional back-propagation NN, FDA, and PCA algorithms.

PCA plus LDA on Wavelet Co-occurrence Histogram Features for Texture Classification and its Applications

paper, we propose a combined approach, namely, PCA plus LDA on Wavelet Co-occurrence Histogram Features (WCHF) for texture classification. The texture features are extracted using the Wavelet Co-occurrence Histogram (WCH) from wavelet decomposed images, which capture the information about relationships between each high frequency subband and that in low frequency subband of the wavelet transformed image at the corresponding level. The correlation between the subbands at the same resolution exhibits a strong relationship, indicating that this information is significant for characterizing a texture. Thus WCH features thus extracted form a feature vector of dimension 384 for gray scale image which is very high. A combination of Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) is applied on WCH feature vector for dimensionality reduction and enhancement of the class separability respectively. The vectors obtained from the LDA are representative of each image. The classification performance is tested on a set of 32 Brodatz textures. The results are compared with the method proposed in (Hiremath and Shivashankar, 2008). The effectiveness of proposed method is demonstrated for two different applications, i.e., CBIR and script identification (both printed and handwritten). The classification performance is analyzed using the k-NN classifier. It is evident from the experimental results that the proposed method exhibits superior performance in the reduced feature set. Keywordsclassification; PCA and LDA; Wavelet; CBIR; document image; script identification

Sexual dimorphism in the feeding mechanism of threespine stickleback

Sexual dimorphism is common in nature and has the potential to increase intraspecific variation in performance and patterns of resource use. We sought to determine whether anadromous threespine stickleback, Gasterosteus aculeatus, exhibit sexual dimorphism in feeding kinematics. We filmed four males and four females consuming live prey in a total of 51 sequences filmed at 500 Hz, then tested for differences in cranial kinematics using a combination of principal component analysis and linear mixed models. We document, for the first time in fishes, divergence between males and females in both the timing of key movements and the magnitude of excursions reached by the hyoid, jaws and neurocranium during prey capture. Some of the largest differences are in jaw protrusion, with males exhibiting faster time to peak jaw protrusion but females exhibiting greater maximum jaw protrusion. Measurements of morphological jaw protrusion on cleared and stained specimens significantly predict jaw protrusion in kinematics. This morphological divergence could reflect ecological divergence between the sexes, or the demands of nest building and territory defense compromising male feeding performance. Remarkably, the morphological jaw protrusion divergence in anadromous males and females is similar to jaw protrusion divergence between ecomorphs in a benthic-limnetic species pair, with limnetics exhibiting female-like patterns of protrusion and benthics exhibiting male-like patterns. These results suggest that sexual dimorphism in feeding functional morphology exists in nature and may have played an important role in the radiation of threespine stickleback.

A rapid and simple approach to discriminate various extracts of Saraca asoca [Roxb.], De. Wild using UPLC-QTOFMS and multivariate analysis

BackgroundBark of Saraca asoca is clinically proved female tonic being used from centuries to treat excessive loss of blood during menstruation in presence of leucorrhea, uterine fibroids and other reasons. However, due to growing demand, all of its parts are in use as substituent instead of bark. MethodsIn the present report, ultra performance liquid chromatography coupled to quadrupole-time-of-flight mass spectrometer was used to distinguish and validate various S. asoca extracts and its herbal formulation ‘Ashokarista’. Principal component analysis and partial least square discriminant analysis [PLS-DA] were used to discriminate the samples. Results & discussionPrincipal component analysis of data showed a clear and significant separation of compositions among the plant parts and drugs. A significant accuracy of developed PLS-DA model for sample validation was found to be 93.333%. The tentative identification of biomarkers using intensity threshold 20,000 counts per second is complemented by the accurate mass measurement of quadrupole-time-of-flight mass spectrometer. The proposed analytical method can be used for authentication of herbal samples. ConclusionThe results demonstrated that the combinations of principal component analysis and PLS-DA offer an efficient way to identify and evaluate different extract of S. asoca. Developed method and identified biomarkers can be used for quality control purpose of S. asoca based herbal drugs.

Regional and local scale atmospheric forcing upon sea level along the coast of SW Europe

ABSTRACTSea level height variability along the coast of southwestern Europe is analyzed in relation with local and regional scale atmospheric forcing mechanisms. A circulation type (CP) catalogue, derived from a combination of principal component analysis and cluster analysis was used to characterize the regional scale atmospheric circulation, whereas the local scale was highlighted through the analysis of hourly wind records during extreme events. A large portion of common anomalous sea level variability at 6‐hourly time frame can be explained by the effect of one CP, depicting a deep extra‐tropical disturbance, close to the Iberian Peninsula, which represents the dominance of the inverse barometer mechanism. Other CPs can induce occasionally similar effects, but showing a spatially restricted scope, in relationship with the location of the main centres of action. Embedded into this synoptic framework, a local forcing, related to both meteorological processes (wind set up) and coastal oceanographic features (depth of the continental shelf) usually control the timing and less frequently, the magnitude of the surges. The decreasing trend observed in the magnitude of the sea level residuals derived from a hindcast model as well as some spatial differences (more intense trend in the upper percentiles along the northern Spain and Mediterranean coasts) can also be explained by long‐term changes in the frequency, and specially, in the internal variability of the CPs, which might be linked to the recent positive phase and eastward displacement of the southern node of the North Atlantic Oscillation. An overall increase in sea level pressure, detected in all CPs, has been accompanied by significant increases in types characterized by anticyclones over the Iberian Peninsula and decreases of northerly flows, related to a reduction of the number of cyclones passing through the Gulf of Biscay and the Western Mediterranean Basin.

Quality aspects of Dutch general practice–based data: a conceptual approach

General practice-based data, collected within general practice registration networks (GPRNs), are widely used in research. The quality of the data is important but the recording criteria about what type of information is collected and how this information should be recorded differ between GPRNs. We aim to identify aspects that describe the quality of general practice-based data in the Netherlands. To investigate the quality aspects, we used the method of concept mapping, a structured conceptualization process for a complex multi-dimensional topic. We explored the ideas of representatives from 10 Dutch GPRNs on the quality of general practice-based data in five steps: preparation, generation of statements, structuring, representation and interpretation. In a brainstorm session, 10 experts generated statements about good data quality from general practice, which we completed with information from the literature. In total, 18 experts participated in the ranking and clustering of the statements. These results were analysed using ARIADNE software, using a combination of principal component analysis and cluster analysis techniques. Finally, the clusters were labelled based on their content. A total of 72 statements were analysed, which resulted in a two-dimensional picture with six clusters, 'complete health record', 'coding of information', 'episode oriented recording', 'diagnostic validity', 'recording agreements' and 'residual category'. The quality of general practice-based data can be considered on five content-based aspects. These aspects determine the quality of recording.

Precise and Accurate Job Cycle Time Forecasting in a Wafer Fabrication Factory with a Fuzzy Data Mining Approach

Many data mining methods have been proposed to improve the precision and accuracy of job cycle time forecasts for wafer fabrication factories. This study presents a fuzzy data mining approach based on an innovative fuzzy backpropagation network (FBPN) that determines the lower and upper bounds of the job cycle time. Forecasting accuracy is also significantly improved by a combination of principal component analysis (PCA), fuzzy c-means (FCM), and FBPN. An applied case that uses data collected from a wafer fabrication factory illustrates this fuzzy data mining approach. For this applied case, the proposed methodology performs better than six existing data mining approaches.

Functional clustering for Italian climate zones identification

This work presents a functional clustering procedure applied to meteorological time series. Our proposal combines time series interpolation with smoothing penalized B-spline and the partitioning around medoids clustering algorithm. Our final goal is to obtain homogeneous climate zones of Italy. We compare this approach to standard methods based on a combination of principal component analysis and Cluster Analysis (CA) and we discuss it in relation to other functional clustering approaches based on Fourier analysis and CA. We show that a functional approach is simpler than the standard methods from a methodological and interpretability point of view. Indeed, it becomes natural to find a clear connection between mathematical results and physical variability mechanisms. We discuss how the choice of the basis expansion (splines, Fourier) affects the analysis and propose some comments on their use. The basis for classification is formed by monthly values of temperature and precipitation recorded during the period 1971–2000 over 95 and 94 Italian monitoring stations, respectively. An assessment based on climatic patterns is presented to prove the consistency of the clustering and a comparison of results obtained with different methods is used to judge the functional data approach.

Model independent early expansion history and dark energy

We examine model independent constraints on the high redshift and prerecombination expansion history from upcoming cosmic microwave background observations, using a combination of principal component analysis and other techniques. This can be translated to model independent limits on early dark energy and the number of relativistic species ${N}_{\mathrm{eff}}$. Models such as scaling (Doran-Robbers), dark radiation ($\ensuremath{\Delta}{N}_{\mathrm{eff}}$), and barotropic aether fall into distinct regions of eigenspace and can be easily distinguished from each other. Incoming CMB data will map the expansion history from $z=0--{10}^{5}$, achieving subpercent precision around recombination, and enable determination of the amount of early dark energy and valuable guidance to its nature.

Correspondence between genetic structure and farmers' taxonomy – a case study from dry-season sorghum landraces in northern Cameroon

The study of the genetic structure of cultivated plant populations maintained by farmers is of great importance for evolutionary and conservation biology. Such studies help understand the bases of crop evolution and conservation in relation to farmers' practices. In this study, we assessed the genetic structure underlying landrace diversity of dry-season sorghum. This crop constitutes a historical model of innovation developed by farmers to extend sorghum cultivation to the dry season. Two types of dry-season sorghum are cultivated. We aimed to assess the link between farmers' taxonomy and molecular genetic structure. We collected both types of dry-season sorghum in two villages of northern Cameroon which represented 20 landraces. These landraces were genotyped using eight polymorphic microsatellite markers. This study compared two clustering methods: a Bayesian method (STRUCTURE) which is based on explicit genetic assumptions and the discriminant analysis of principal component method. The latter, more recently proposed, is based on the combination of principal component analysis and discriminant analysis. We noticed a general congruence between these two methods. We also used both methods to infer the genetic structure of our sample. Our results showed strong genetic structuring of the landraces, withK= 14 genetic clusters. We then analysed the fit between farmers' taxonomy and genetic structure. The data suggested that each type and each landrace corresponds to a given genetic entity. This pattern was robust across both villages, despite differences in cultural practices.

Optical Fiber Sensors Array to Identify Beverages by Their Odor

Four optical fiber sensors have been grouped in an array which is able to distinguish odors of different drinks. The sensing materials employed have been deposited onto optical fibers following the electrostatic self assembly method. The responses have been characterized in terms of reflected optical power; more specifically, the dynamic range and the recovery of each device have been used to discriminate between the samples. Data mining techniques based on the combination of principal component analysis and artificial neural networks are performed. The final system is trained to distinguish between grape juice, wine, and vinegar by using a set of one hundred samples of each one. Furthermore, the array can be located at up to 6 km away from the optical header, offering the possibility of in situ measurements.

Sensing, Compression, and Recovery for WSNs: Sparse Signal Modeling and Monitoring Framework

We address the problem of compressing large and distributed signals monitored by a Wireless Sensor Network (WSN) and recovering them through the collection of a small number of samples. We propose a sparsity model that allows the use of Compressive Sensing (CS) for the online recovery of large data sets in real WSN scenarios, exploiting Principal Component Analysis (PCA) to capture the spatial and temporal characteristics of real signals. Bayesian analysis is utilized to approximate the statistical distribution of the principal components and to show that the Laplacian distribution provides an accurate representation of the statistics of real data. This combined CS and PCA technique is subsequently integrated into a novel framework, namely, SCoRe1: Sensing, Compression and Recovery through ON-line Estimation for WSNs. SCoRe1 is able to effectively self-adapt to unpredictable changes in the signal statistics thanks to a feedback control loop that estimates, in real time, the signal reconstruction error. We also propose an extensive validation of the framework used in conjunction with CS as well as with standard interpolation techniques, testing its performance for real world signals. The results in this paper have the merit of shedding new light on the performance limits of CS when used as a recovery tool in WSNs.

Computing Integrated Ratings from Heterogeneous Phenotypic Assessments: A Case Study of Lettuce Postharvest Quality and Downy Mildew Resistance

ABSTRACTComparing performance of a large number of accessions simultaneously is not always possible. Frequently, only subsets of all accessions are tested in separate trials with only some (or none) of the accessions overlapping between subsets. Using standard statistical approaches to combine data from such a sparsely populated accession × trial matrix is precluded if different rating scales are used to evaluate accessions in those trials. Here we compare two approaches that can compute an overall linear rating for the performance of all accessions across a set of trials, even for accessions that were never tested together and were rated on dissimilar scales. We use data from lettuce (Lactuca sativa L.) postharvest quality collected on 178 accessions in 18 trials and assessment of lettuce resistance to downy mildew (Bremia lactucae Regel) performed on 583 accessions in 53 trials. The projected values (PV) approach uses a combination of principal component analysis and resampling to merge trial results and calculates an overall rating from real values. In contrast, the rank‐aggregation (RA) approach uses an extension of the Rasch model to combine rank‐ordered data from individual trials. We found high correlation between ratings produced by the two approaches for the postharvest quality (r= 0.803) and the resistance to downy mildew (r = 0.748). Combining data from multiple experiments identified lettuce accessions with a high level of resistance to the disease and a slow rate of deterioration when processed for salad. The PV and RA approaches also allow combining data from different laboratories or databases.

Comparison of analytical and sensory lipid oxidation parameters in conventional and high‐oleic rapeseed oil

AbstractHeadspace‐solid‐phase microextraction‐gas chromatography was used to identify in total 55 volatile lipid oxidation compounds in thermally stressed conventional and high‐oleic rapeseed oil samples. Out of this profile, 17 volatile compounds with low odor threshold values were selected as target compounds for the assessment of lipid oxidation in rapeseed and high‐oleic rapeseed oils. Additionally, other lipid oxidation parameters such as fatty acid composition, peroxide value, anisidine value, and induction time (Rancimat analysis) were determined. Multivariate statistical methods (principal component analysis in combination with agglomerative hierarchical cluster analysis) were applied to identify sensitive volatile lipid oxidation indicators enabling the differentiation of rapeseed oil samples of different varieties (high‐oleic versus conventional). Moreover, these statistical methods were capable of differentiating rapeseed oils of different oxidative properties. Octanal and 3‐octanone showed the highest ability to differentiate between samples of different rapeseed varieties, whereas propanal, E,E‐2,4‐hexadienal, and E‐2‐heptenal were most suitable in differentiating rapeseed oil samples with different oxidative properties from each other. Clustering of rapeseed oil samples according to their volatile compound composition was comparable with results of sensory duo‐trio and paired comparison tests, but the analytical approach of the volatile compound analysis in combination with chemometric methods detected changes sooner in relation to the flavor composition of rapeseed oils and high‐oleic rapeseed oil samples.Practical applications:The combination of volatile compound analysis by HS‐SPME‐GC with multivariate statistical methods and complementary sensory duo‐trio and one‐sided paired comparison tests are sensitive tools in differentiating conventional and high‐oleic rapeseed oil samples with different lipid oxidation properties. The presented methods are suitable techniques for the detection of initial changes of lipid oxidation progress in edible oils.

An Approach for ECG Characterization and Classification Using the Combination of Wavelet Transform and Decision Tree Methods

A new automated approach for Electrocardiogram (ECG) arrhythmias characterization and classification with the combination of Wavelet transform and Decision tree classification is presented. The approach is based on two key steps. In the first step, the authors adopt the wavelet transform to extract the ECG signals wavelet coefficients as first features and utilize the combination of Principal Component Analysis (PCA) and Fast Independent Component Analysis (FastICA) to transform the first features into uncorrelated and mutually independent new features. In the second step, they utilize some decision tree methods currently in use: C4.5, Improved C4.5, CHAID (Chi - Square Automatic Interaction Detection) and Improved CHAID for the classification of ECG signals, which are taken, from the MIT-BIH database, including normal subjects and subjects affected by arrhythmia. The authors’ results suggest the high reliability and high classification accuracy of C4.5 algorithm with the bootstrap aggregation.

An Affective Approach to Developing Marketing Strategies of Mineral Water

This study is concerned with the development of marketing strategies for mineral water based on consumers’ taste preferences, by analyzing the taste components of mineral water. In this study, we used a twodimensional analysis to classify taste data. We conducted a correlation analysis to identify the characteristics of taste data. We applied a combination of principal component analysis and self-organizing map to classify mineral water tastes. Based on this evaluation, we identified some marketing strategies in the conclusion. According to this study, the taste of mineral water is not determined by the origin and is not influenced by the hardness of the water.

Special Issue on Modeling and Simulation

Special Issue on Modeling and Simulation