Performance Of Principal Component Analysis Research Articles

Evaluating the Effect of the Financial Status to the Mobility Customs

In this article, we explore the relationship between cellular phone data and housing prices in Budapest, Hungary. We determine mobility indicators from one months of Call Detail Records (CDR) data, while the property price data are used to characterize the socioeconomic status at the Capital of Hungary. First, we validated the proposed methodology by comparing the Home and Work locations estimation and the commuting patterns derived from the cellular network dataset with reports of the national mini census. We investigated the statistical relationships between mobile phone indicators, such as Radius of Gyration, the distance between Home and Work locations or the Entropy of visited cells, and measures of economic status based on housing prices. Our findings show that the mobility correlates significantly with the socioeconomic status. We performed Principal Component Analysis (PCA) on combined vectors of mobility indicators in order to characterize the dependence of mobility habits on socioeconomic status. The results of the PCA investigation showed remarkable correlation of housing prices and mobility customs.

Ageratina adenophora and Lantana camara in Kailash Sacred Landscape, India: Current distribution and future climatic scenarios through modeling.

The Himalayan region is one of the global biodiversity hotspots. However, its biodiversity and ecosystems are threatened due to abiotic and biotic drivers. One of the major biotic threats to biodiversity in this region is the rapid spread of Invasive Alien Species (IAS). Natural forests and grasslands are increasingly getting infested by IAS affecting regeneration of native species and decline in availability of bio-resources. Assessing the current status of IAS and prediction of their future spread would be vital for evolving specific species management interventions. Keeping this in view, we conducted an in-depth study on two IASs, viz., Ageratina adenophora and Lantana camara in the Indian part of Kailash Sacred Landscape (KSL), Western Himalaya. Intensive field surveys were conducted to collect the presence of A. adenophora (n = 567) and L. camara (n = 120) along an altitudinal gradient between 300 and 3000 m a.s.l. We performed Principal Component Analysis to nullify the multi-colinearity effects of the environmental predictors following MaxEnt species distribution model in the current and future climatic scenarios for both the species. All current and future model precision (i.e., Area Under the Curve; AUC) for both species was higher than 0.81. It is predicted that under the current rate of climate change and higher emission (i.e., RCP 8.5 pathway), A. adenophora will spread 45.3% more than its current distribution and is likely to reach up to 3029 m a.s.l., whereas, L. camara will spread 29.8% more than its current distribution range and likely to reach up to 3018 m a.s.l. Our results will help in future conservation planning and participatory management of forests and grasslands in the Kailash Sacred Landscape–India.

Recognition of Augmented Frontal Face Images Using FFT-PCA/SVD Algorithm

In spite of the differences in visual stimulus of human beings such as ageing, changing conditions of a person, and occlusion, recognition can even be done at a glance by the human eye many years after the previous encounter. It has been established that facial differences like the hairstyle changes, growing of one’s beard, wearing of glasses, and other forms of occlusions can hardly hinder the power of the human brain from making a face recognition. However, the same cannot easily be said about automated intelligent systems which have been developed to mimic the skill of the human brain to aid in recognition. There have been growing interests in developing a resilient and efficient recognition system mainly because of its numerous application areas (access control, entertainment/leisure, security system based on biometric data, and user-friendly human-machine interfaces). Although there have been numerous researches on face recognition under varying pose, illumination, expression, and image degradations, problems caused by occlusions are mostly ignored. This study thus focuses on facial occlusions and proposes an enhancement mechanism through face image augmentation to improve the recognition of occluded face images. This study assessed the performance of Principal Component Analysis with Singular Value Decomposition using Fast Fourier Transform (FFT-PCA/SVD) for preprocessing face recognition algorithm on face images with missingness and augmented face image database. It was found that the average recognition rates for the FFT-PCA/SVD algorithm were the same ( 90 % ) when face images with missingness and augmented face images were used as test images, respectively. The statistical evaluation revealed that there exists a significant difference in the average recognition distances for the face images with missingness and augmented face images when FFT-PCA/SVD is used for recognition. Augmented face images tend to have a relatively lower average recognition distance when used as test images. This finding is contrary to the equal performance assessment by the adopted numerical technique. The MICE algorithm is therefore recommended as a suitable imputation mechanism for enhancing/improving the performance of the face recognition system.

Cerebral Amyloid Angiopathy Burden and Cerebral Microbleeds: Pathological Evidence for Distinct Phenotypes.

The relationship between cerebral microbleeds (CMBs) on hemosiderin-sensitive MRI sequences and cerebral amyloid angiopathy (CAA) remains unclear in population-based participants or in individuals with dementia. To determine whether CMBs on antemortem MRI correlate with CAA. We reviewed 54 consecutive participants with antemortem T2*GRE-MRI sequences and subsequent autopsy. CMBs were quantified on MRIs closest to death. Autopsy CAA burden was quantified in each region including leptomeningeal/cortical and capillary CAA. By a clustering approach, we examined the relationship among CAA variables and performed principal component analysis (PCA) for dimension reduction to produce two scores from these 15 interrelated predictors. Hurdle models assessed relationships between principal components and lobar CMBs. MRI-based CMBs appeared in 20/54 (37%). 10 participants had ≥2 lobar-only CMBs. The first two components of the PCA analysis of the CAA variables explained 74% variability. The first rotated component (RPC1) consisted of leptomeningeal and cortical CAA and the second rotated component of capillary CAA (RPC2). Both the leptomeningeal and cortical component and the capillary component correlated with lobar-only CMBs. The capillary CAA component outperformed the leptomeningeal and cortical CAA component in predicting lobar CMBs. Both capillary and the leptomeningeal/cortical components correlated with number of lobar CMBs. Capillary and leptomeningeal/cortical scores correlated with lobar CMBs on MRI but lobar CMBs were more closely associated with the capillary component. The capillary component correlated with APOEɛ4, highlighting lobar CMBs as one aspect of CAA phenotypic diversity. More CMBs also increase the probability of underlying CAA.

Fast Strain Estimation and Frame Selection in Ultrasound Elastography Using Machine Learning.

Ultrasound elastography aims to determine the mechanical properties of the tissue by monitoring tissue deformation due to internal or external forces. Tissue deformations are estimated from ultrasound radio frequency (RF) signals and are often referred to as time delay estimation (TDE). Given two RF frames I1 and I2 , we can compute a displacement image, which shows the change in the position of each sample in I1 to a new position in I2 . Two important challenges in TDE include high computational complexity and the difficulty in choosing suitable RF frames. Selecting suitable frames is of high importance because many pairs of RF frames either do not have acceptable deformation for extracting informative strain images or are decorrelated and deformation cannot be reliably estimated. Herein, we introduce a method that learns 12 displacement modes in quasi-static elastography by performing principal component analysis (PCA) on displacement fields of a large training database. In the inference stage, we use dynamic programming (DP) to compute an initial displacement estimate of around 1% of the samples and then decompose this sparse displacement into a linear combination of the 12 displacement modes. Our method assumes that the displacement of the whole image could also be described by this linear combination of principal components. We then use the GLobal Ultrasound Elastography (GLUE) method to fine-tune the result yielding the exact displacement image. Our method, which we call PCA-GLUE, is more than 10× faster than DP in calculating the initial displacement map while giving the same result. This is due to converting the problem of estimating millions of variables in DP into a much simpler problem of only 12 unknown weights of the principal components. Our second contribution in this article is determining the suitability of the frame pair I1 and I2 for strain estimation, which we achieve by using the weight vector that we calculated for PCA-GLUE as an input to a multilayer perceptron (MLP) classifier. We validate PCA-GLUE using simulation, phantom, and in vivo data. Our classifier takes only 1.5 ms during the testing phase and has an F1-measure of more than 92% when tested on 1430 instances collected from both phantom and in vivo data sets.

Construction of a SNP Fingerprinting Database and Population Genetic Analysis of Cigar Tobacco Germplasm Resources in China.

Cigar tobacco is an important economic crop that is widely grown around the world. In recent years, varietal identification has become a frequent problem in germplasm preservation collections, which causes considerable inconvenience and uncertainty in the cataloging and preservation of cigar germplasm resources, in the selection of parental lines for breeding, and in the promotion and use of high quality varieties. Therefore, the use of DNA fingerprints to achieve rapid and accurate identification of varieties can play an important role in germplasm identification and property rights disputes. In this study, we used genotyping-by-sequencing (GBS) on 113 cigar tobacco accessions to develop SNP markers. After filtering, 580,942 high-quality SNPs were obtained. We used the 580,942 SNPs to perform principal component analysis (PCA), population structure analysis, and neighbor joining (NJ) cluster analysis on the 113 cigar tobacco accessions. The results showed that the accessions were not completely classified based on their geographical origins, and the genetic backgrounds of these cigar resources are complex and diverse. We further selected from these high-quality SNPs to obtained 163 SNP sites, 133 of which were successfully converted into KASP markers. Finally, 47 core KASP markers and 24 candidate core markers were developed. Using the core markers, we performed variety identification and fingerprinting in 216 cigar germplasm accessions. The results of SNP fingerprinting, 2D barcoding, and genetic analysis of cigar tobacco germplasm in this study provide a scientific basis for screening and identifying high-quality cigar tobacco germplasm, mining important genes, and broadening the basis of cigar tobacco genetics and subsequent breeding work at the molecular level.

RAMP2-AS1 Regulates Endothelial Homeostasis and Aging.

The homeostasis of vascular endothelium is crucial for cardiovascular health and endothelial cell (EC) aging and dysfunction could negatively impact vascular function. Leveraging transcriptome profiles from ECs subjected to various stimuli, including time-series data obtained from ECs under physiological pulsatile flow vs. pathophysiological oscillatory flow, we performed principal component analysis (PCA) to identify key genes contributing to divergent transcriptional states of ECs. Through bioinformatics analysis, we identified that a long non-coding RNA (lncRNA) RAMP2-AS1 encoded on the antisense of RAMP2, a determinant of endothelial homeostasis and vascular integrity, is a novel regulator essential for EC homeostasis and function. Knockdown of RAMP2-AS1 suppressed RAMP2 expression and caused EC functional changes promoting aging, including impaired angiogenesis and increased senescence. Our study demonstrates an integrative approach to quantifying EC aging based on transcriptome changes, which also identified a number of novel regulators, including protein-coding genes and many lncRNAs involved EC functional modulation, exemplified by RAMP2-AS1.

Sparse principal component regression via singular value decomposition approach

Principal component regression (PCR) is a two-stage procedure: the first stage performs principal component analysis (PCA) and the second stage builds a regression model whose explanatory variables are the principal components obtained in the first stage. Since PCA is performed using only explanatory variables, the principal components have no information about the response variable. To address this problem, we present a one-stage procedure for PCR based on a singular value decomposition approach. Our approach is based upon two loss functions, which are a regression loss and a PCA loss from the singular value decomposition, with sparse regularization. The proposed method enables us to obtain principal component loadings that include information about both explanatory variables and a response variable. An estimation algorithm is developed by using the alternating direction method of multipliers. We conduct numerical studies to show the effectiveness of the proposed method.

Development and validation of a structured feedback questionnaire from postgraduates on various elements of postgraduate medical curriculum

BackgroundMedical Council of India, introduced the Post Graduate (PG) curriculum as ‘Competency Based Medical Education’ (CBME). Feedback from the end users is a vital step in curriculum evaluation. Therefore, the primary objective of this study was to develop and validate a Structured Feedback Questionnaire (SFQ) for postgraduates, encompassing all the components of the PG-CBME curriculum. MethodsSFQ was developed with 23 Likert based questions and four open ended questions. Content validation was done by Lawshe method. After getting institutional ethics clearance and informed consent, SFQ was administered to 121 final year PGs (response rate 100%). We performed Principal component analysis (PCA), Structural equation modeling (SEM), Chi squared test (χ2/df); goodness-of-fit index (GFI); adjusted GFI; comparative fit index (CFI) and root mean square error of approximation (RMSEA). Cronbach's alpha was done for estimating the internal consistency. ResultsThe validation resulted in a three-factor model comprising of “curriculum” (42.1%), “assessment” (28%), and “support” (18.5%). Chi squared test (χ2/df ratio) < 2, CFI (0.78), GFI (0.72) and RMSEA (0.09) indicated superior goodness of fit for the three-factor model for the sample data. All the extracted factors had good internal consistency of ≥0.9. ConclusionWe believe that this 23 item SFQ is a valid and reliable tool which can be utilized for curriculum evaluation and thereby formulating recommendations to modify the existing curriculum wherever required, facilitating enriched program outcomes.

Univariate and Multivariate QTL Analyses Reveal Covariance Among Mineral Elements in the Rice Ionome.

Rice provides more than one fifth of daily calories for half of the world’s human population, and is a major dietary source of both essential mineral nutrients and toxic elements. Rice grains are generally poor in some essential nutrients but may contain unsafe levels of some toxic elements under certain conditions. Identification of quantitative trait loci (QTLs) controlling the concentrations of mineral nutrients and toxic trace metals (the ionome) in rice will facilitate development of nutritionally improved rice varieties. However, QTL analyses have traditionally considered each element separately without considering their interrelatedness. In this study, we performed principal component analysis (PCA) and multivariate QTL analyses to identify the genetic loci controlling the covariance among mineral elements in the rice ionome. We resequenced the whole genomes of a rice recombinant inbred line (RIL) population, and performed univariate and multivariate QTL analyses for the concentrations of 16 elements in grains, shoots and roots of the RIL population grown in different conditions. We identified a total of 167 unique elemental QTLs based on analyses of individual elemental concentrations as separate traits, 53 QTLs controlling covariance among elemental concentrations within a single environment/tissue (PC-QTLs), and 152 QTLs which determined covariation among elements across environments/tissues (aPC-QTLs). The candidate genes underlying the QTL clusters with elemental QTLs, PC-QTLs and aPC-QTLs co-localized were identified, including OsHMA4 and OsNRAMP5. The identification of both elemental QTLs and PC QTLs will facilitate the cloning of underlying causal genes and the dissection of the complex regulation of the ionome in rice.

How Selective Breeding Has Changed the Morphology of the American Mink (Neovison vison)-A Comparative Analysis of Farm and Feral Animals.

Simple SummaryDecades of selective breeding carried out on fur farms have changed the morphology, behavior and other features of the American mink, thereby differentiating farm and feral animals. The uniqueness of this situation is not only that we can observe how selective breeding phenotypically and genetically changes successive generations, but also that it enables a comparison of farm minks with their feral counterparts. Such a comparison may thus provide valuable information regarding differences in natural selection and selective breeding. In our study, we found significant morphological differences between farm and feral minks as well as changes in body shape: trapezoidal in feral minks and rectangular in farm minks. Such a clear differentiation between the two populations over a period of several decades highlights the intensity of selective breeding in shaping the morphology of these animals and gives an indication of the speed of phenotypic changes and the species’ plasticity. This also suggests that the selective forces (selective breeding vs. natural selection) acting upon body dimensions of minks vary between feral and farm populations.In this study, we performed a comparative analysis of the morphological traits between feral (n = 43) and farm (n = 200) individuals of the American mink in Poland to address the question of how multigenerational intensive selective breeding has morphologically differentiated these two populations. Nine body measurements and two proportion coefficients were obtained using adult individuals. The significance of differences between population means was assessed using the Wilcoxon test for independent samples, while the Kruskal–Wallis test was used to compare sex-population groups. Spearman’s correlation coefficients between measurements were estimated for each population. We also performed Principal Component Analysis (PCA) to identify the variables that were most closely correlated with variation in the trait measurements and to investigate the morphological differences between farm and feral minks. We found that the farm minks exhibited significantly higher mean values for eight out of eleven studied traits. Moreover, significant changes in forelimb length, with no concomitant changes in hindlimb length, were accompanied by differences in body shape: trapezoidal in feral minks and rectangular in farm minks. The PCA suggested an almost complete separation of the two populations and indicated that sexes were quite separate; farm males in particular constitute a wholly discrete cluster. Such a clear differentiation between the two populations and sexes over a period of several decades highlights the intensity of selective breeding in shaping the morphology of these animals.

Large deviations of extreme eigenvalues of generalized sample covariance matrices

We present an analytical technique to compute the probability of rare events in which the largest eigenvalue of a random matrix is atypically large (i.e., the right tail of its large deviations). The results also transfer to the left tail of the large deviations of the smallest eigenvalue. The technique improves upon past methods by not requiring the explicit law of the eigenvalues, and we apply it to a large class of random matrices that were previously out of reach. In particular, we solve an open problem related to the performance of principal components analysis on highly correlated data, and open the way towards analyzing the high-dimensional landscapes of complex inference models. We probe our results using an importance sampling approach, effectively simulating events with probability as small as .

Beyond the Mini-Mental State Examination: The Use of Physical and Spatial Navigation Tests to Help to Screen for Mild Cognitive Impairment and Alzheimer's Disease.

Spatial navigation and dual-task (DT) performance may represent a low-cost approach to the identification of the cognitive decline in older adults and may support the clinical diagnosis of mild cognitive impairment (MCI) and Alzheimer's disease (AD). To assess the accuracy of different types of motor tasks in differentiating older persons with MCI and AD from healthy peers. Older adults aged 60 years or over (n = 105; healthy = 39; MCI = 23; AD = 43) were evaluated by the floor maze test (FMT), the senior fitness test, and DT performance. Receiver operating characteristic curve (ROC) analysis was used to evaluate the accuracy of the tests. We also performed principal component analysis (PCA) and logistic regression analysis to explore the variance and possible associations of the variables within the sample. FMT (AUC = 0.84, sensitivity = 75.7%, specificity = 76.1%, p < 0.001) and DT (AUC = 0.87, sensitivity = 80.4%, specificity = 86.9%, p < 0.001) showed the highest performance for distinguishing MCI from AD individuals. Moreover, FMT presented better sensitivity in distinguishing AD patients from their healthy peers (AUC = 0.93, sensitivity = 94%, specificity = 85.6%, p < 0.001) when compared to the Mini-Mental State Examination. PCA revealed that the motor test performance explains a total of 73.9% of the variance of the sample. Additionally, the results of the motor tests were not influenced by age and education. Spatial navigation tests showed better accuracy than usual cognitive screening tests in distinguishing patients with neurocognitive disorders.

Greedy Broad Learning System

In order to overcome the extremely time-consuming drawback of deep learning (DL), broad learning system (BLS) was proposed as an alternative method. This model is simple, fast, and easy to update. To ensure the fitting and generalization ability of BLS, the hidden layer neurons are often set too many, in fact, a lot of neurons are not needed. Greedy BLS (GBLS) is proposed in this paper to deal with the redundancy of the hidden layer in BLS from another perspective. Different from BLS, the structure of GBLS can be seen as a combination of unsupervised multi-layer feature representation and supervised classification or regression. It trains with a greedy learning scheme, performs principal component analysis (PCA) on the previous hidden layer to form a set of compressed nodes, which are transformed into enhancement nodes and then activated by nonlinear functions. The new hidden layer is composed of all newly generated compressed nodes and enhancement nodes, and so on. The last hidden layer of the network contains the higher-order and abstract essential features of the original data, which is connected to the output layer. Each time a new layer is added to the model, and there is no need to retrain from the beginning, only the previous layer is trained. Experimental results demonstrate that the proposed GBLS model outperforms BLS both in classification and regression.

Principal component analysis of routine blood test results with Parkinson's disease: A case-control study.

This study aimed to explore the association of routine blood test values and blood cell ratios with the risk or severity of Parkinson's disease (PD). The medical records of 453 PD patients and 436 controls were retrospectively reviewed. The severity of PD was quantified by the modified Hoehn-Yahr (HY) scale. We performed principal component analysis (PCA) of significant values/ratios and used logistic regression analysis to explore the relationship between principal components (PCs) and the risk of PD. Spearman correlation and ordinal logistic regression analyses were performed to explore the relationship between indicators and the severity of PD. The PCA generated 9 PCs, which contributed to 90.86% of the total variance. Logistic regression analysis revealed positive associations of PC2 (a measure monocyte ratios) and PC6 (a measure of platelet ratios and volume) and negative associations of PC1 (a comprehensive measure of lymphocyte, eosinophil, neutrophil, and red blood cell values), PC4 (a measure of red blood cell values), and PC7 (a measure of red blood cell values and platelet volume) with the risk of PD. However, we observed no associations of variables with the severity of PD. In conclusion, PCA reduced the dimensionality of the data. Peripheral blood disorders may be associated with PD.

402 Audio information retrieval for describing gait patterns in Brazilian horses

Abstract The Brazilian gaited horse industry is growing steadily, even after a recession period that affected different economic sectors in the whole country. Recent numbers suggested an increase on the exports, which reveals the relevance of this horse market segment. Horses are classified according to the gait criteria, which divide the horses in two groups associated with the animal movements: lateral (Marcha Picada) or diagonal (Marcha_Batida). These two gait groups usually show remarkable differences related to speed and number of steps per fixed unit of time, among other factors. Audio retrieval refers to the process of information extraction obtained from audio signals. This new data analysis area, in comparison to traditional methods to evaluate and classify gait types (as, for example, human subjective evaluation and video monitoring), provides a potential method to collect phenotypes in a reduced cost manner. Audio files (n = 80) were obtained after extracting audio features from freely available YouTube videos. Videos were manually labeled according to the two gait groups (Marcha Picada or Marcha Batida) and thirty animals were used after a quality control filter step. This study aimed to investigate different metrics associated with audio signal processing, in order to first cluster animals according to the gait type and subsequently include additional traits that could be useful to improve accuracy during the identification of genetically superior animals. Twenty-eight metrics, based on frequency or physical audio aspects, were carried out individually or in groups of relative importance to perform Principal Component Analysis (PCA), as well as to describe the two gait types. The PCA results indicated that over 87% of the animals were correctly clustered. Challenges regarding environmental interferences and noises must be further investigated. These first findings suggest that audio information retrieval could potentially be implemented in animal breeding programs, aiming to improve horse gait.

A Comparative Assessment of the Baking Quality of Hybrid and Population Wheat Cultivars

The study assessed the quality parameters of grain and flour, the rheological properties of dough and the quality of bread prepared from flour of hybrid cultivars of wheat in comparison with population cultivars of wheat. As the interest in wheat hybrids cultivars from the agricultural and milling industry is growing, their technological value of grain and flour was evaluated at two levels of nitrogen fertilisation (N1—110 kg/ha, N2—150 kg/ha). Increasing the fertilisation (N2) produced a significant influence on the crude protein and gluten content in the flour, as well as the moisture of the crumb and the yield of the dough without impacting other rheological traits and parameters of bread baking process. The performed principal component analysis (PCA) allowed for identification of the best cultivars among the studied wheat cultivars (Hybery and Hyvento). The hybrid cultivar Hyvento was characterised by favourable qualitative traits of the grain (vitreousness, crude protein content) and rheological parameters of the dough (bread volume), however, it had lower baking quality parameters. Among the hybrid cultivars, the best applicability for baking purposes was Hybery due to the favourable values of the baking process parameters and bread quality (bread yield, bread volume, Dallmann porosity index of crumb). Hybrid cultivars of wheat can therefore be used for the production of bread and be an alternative in agricultural production for population cultivars, which will contribute to filling the knowledge gap for the hybrid wheat cultivars.

Recognition of Reconstructed Frontal Face Images Using FFT-PCA/SVD Algorithm

Face recognition has gained prominence among the various biometric-based methods (such as fingerprint and iris) due to its noninvasive characteristics. Modern face recognition modules/algorithms have been successful in many application areas (access control, entertainment/leisure, security system based on biometric data, and user-friendly human-machine interfaces). In spite of these achievements, the performance of current face recognition algorithms/modules is still inhibited by varying environmental constraints such as occlusions, expressions, varying poses, illumination, and ageing. This study assessed the performance of Principal Component Analysis with singular value decomposition using Fast Fourier Transform (FFT-PCA/SVD) for preprocessing the face recognition algorithm on left and right reconstructed face images. The study found that average recognition rates for the FFT-PCA/SVD algorithm were 95% and 90% when the left and right reconstructed face images are used as test images, respectively. The result of the paired sample t-test revealed that the average recognition distances for the left and right reconstructed face images are not significantly different when FFT-PCA/SVD is used for recognition. FFT-PCA/SVD is recommended as a viable algorithm for recognition of left and right reconstructed face images.

Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products.

Hyperspectral imaging (HSI) emerges as a non-destructive and rapid analytical tool for assessing food quality, safety, and authenticity. This work aims to investigate the potential of combining the spectral and spatial features of HSI data with the aid of deep learning approach for the pixel-wise classification of food products. We applied two strategies for extracting spatial-spectral features: (1) directly applying three-dimensional convolution neural network (3-D CNN) model; (2) first performing principal component analysis (PCA) and then developing 2-D CNN model from the first few PCs. These two methods were compared in terms of efficiency and accuracy, exemplified through two case studies, i.e., classification of four sweet products and differentiation between white stripe (“myocommata”) and red muscle (“myotome”) pixels on salmon fillets. Results showed that combining spectral-spatial features significantly enhanced the overall accuracy for sweet dataset, compared to partial least square discriminant analysis (PLSDA) and support vector machine (SVM). Results also demonstrated that spectral pre-processing techniques prior to CNN model development can enhance the classification performance. This work will open the door for more research in the area of practical applications in food industry.

Molecular Distributions of the Protostellar Envelope and the Outflow of IRAS 15398–3359: Principal Component Analysis

Abstract Using the Atacama Large Millimeter/submillimeter Array, we have imaged 15 molecular-line emissions and the dust continuum emission around the Class 0 protostellar source IRAS 15398−3359. The outflow structure is mainly traced by the H2CO (K a = 0 and 1), CCH, and CS emissions. These lines also trace the disk/envelope structure around the protostar. The H2CO (K a = 2 and 3), CH3OH, and SO emissions are concentrated toward the protostar, while the DCN emission is more extended around the protostar. We have performed principal component analysis (PCA) for these distributions on two different scales, the outflow and the disk/envelope structure. For the latter case, the molecular-line distributions are classified into two groups, according to the contribution of the second principal component, one having a compact distribution around the protostar and the other showing a rather extended distribution over the envelope. Moreover, the second principal component value tends to increase as an increasing quantum number of H2CO (K a = 0, 1, 2, and 3), reflecting the excitation condition: the distribution is more compact for higher excitation lines. These results indicate that PCA is effective at extracting the characteristic features of the molecular-line distributions around the protostar in an unbiased way. In addition, we identify four blobs in the outflow structure in the H2CO lines, some of which can also be seen in the CH3OH, CS, CCH, and SO emissions. The gas temperature derived from the H2CO lines ranges from 43–63 K, which suggests shocks due to the local impact of the outflow on clumps of the ambient gas.