Multivariate Data Analysis Techniques Research Articles

Evaluation of ORAC, IR and NMR metabolomics for predicting ripening stage and variety in melon (Cucumis melo L.)

A combination of FTIR- and NMR-based metabolomics approaches coupled to multivariate data analysis techniques was used for the first time to characterize and explore the metabolic changes along the ripening of two Spanish melon landraces (Jimbee N and Jimbee XL). Furthermore, the fatty acids profile and the antioxidant capacity were investigated by GC-FID and ORAC method, respectively. Despite FTIR was able to identify changes in the sugars content between fruits of different maturity stages, it failed to discriminate between both varieties, being exceeded by NMR in both situations. Increased fruit maturation led to an increase of sucrose, fatty acids and β-carotene, accompanied by a decrease of glucose, fructose, citrate, amino acids, and polyphenols (which were positively correlated with fruits antioxidant activity).

Symmetrical Robust Canonical Correlation Analysis for Image Classification

Canonical correlation analysis (CCA) is a useful technique for multivariate data analysis, which can find correlations between two sets of multidimensional data. CCA projects two sets of data into a low-dimensional space in which the correlations between them are maximized. However, CCA is sensitive to noise or outliers in the collected data of real-world applications, which will degrade its performance. To overcome this disadvantage, we propose symmetrical robust canonical correlation analysis (SRCCA) for image classification. By using low-rank learning, the noise is removed, and CCA is used to encode correlations between images and their symmetry samples. To verify effectiveness, four public image databases were tested. The result was that SRCCA was more robust than CCA and had good performance for image classification.

Innovation capabilities and the organization of the firm: evidence from Brazil

PurposeThe purpose of this paper is to examine different types of organization of the firm considering the innovation capabilities of manufacturing firms.Design/methodology/approachThe authors carried out an innovation survey with Brazilian manufacturing firms. A sample of 1,156 firms was analyzed in this paper. Collected data were analyzed using multivariate data analysis techniques. From an innovation capabilities approach, it was possible to identify different types of organization of the firm.FindingsResults show four different types of organization of the firm: advanced, intermediate and basic stability-oriented and change-oriented. Each type presents a different innovation capabilities arrangement. The successful strategies toward innovation are related to change-oriented organization of the firm and advanced stability-oriented organization of the firm.Research limitations/implicationsThis study contributes to the literature by presenting a different view on the organization of the firm, encompassing the capabilities approach and thus a higher level on the perception of firms' heterogeneity. This study contributes to narrow the literature gap on how firms internally coordinate its different capabilities into a coherent organization to sustain an innovative behavior.Practical implicationsThese straightforward findings can serve as a guideline so that managers can conduct changes within their companies toward more innovation. Managers can reconsider its organization as a way to foment innovation, once it is identified as a key strategy for competitiveness.Social implicationsThis study may help managers understand that focusing on stability-driven capabilities is riskier if change-driven capabilities are not present in an adequate and aligned level of development. The outcome may be the growth of the cost structure greater than the potential return. Conversely, managers should also understand that once change-driven capabilities are in a glance, they need do follow up with stability-driven capabilities. Here, the risk is not having an adequate structure to sustain the upcoming growth, arising from innovation. In short, not only “cost and value” should be taken together, but they must be arranged following the specific situation of the company. Every company should manage costs either to sustain new added value or to allow the addition of new value.Originality/valueThe study is based on a unique dataset that traces a large set of companies, being able to check different types of firm organization and associate it with innovation capabilities. The study relates to an emerging economy, which has not received adequate attention until now, largely because of the lack of micro-level data. The study is based on a robust theoretical model of innovation capabilities, which is being tested through such data. Finally, results elucidate ways to improve innovation performance of firms.

Influence of Ship Design Complexity on Ship Design Competitiveness

Complexity is discussed in design literature mainly through its negative and in some cases positive consequences. This article critically reviews and elaborates the effects of complexity on competitiveness in ship design, its directionality, and magnitude. The article introduces a model for the measurement of ship design complexity and ship design competitiveness based on predefined factors. Archival data of 100 ship design projects from eight different Norwegian designers are used as case study. Multivariate data analysis techniques are employed to study the research model. The results show a significant correlation between complexity and competitiveness in ship design, where the magnitude and directionality of influence vary among different complexity factors. Our findings provide a basis for enhancing complexity management in ship design. Introduction Continual technology improvements and market volatility with its associated uncertainties have a significant impact on and partly change ship design customers’ expectations. To be successful in such a market, not only does it require the development of competitive products but also the accompanying work processes and the organization or firm framing the development of the vessel solution are involved. Ulstein and Brett (2015) define ship design competitiveness in terms of doing the right thing (effectiveness), doing the right thing right (efficiency), and with the right resources (efficacy) to cover product, process, and firm aspects of competitiveness. To improve their success, ship design companies typically tend to focus on the introduction of new technologies, and, in some cases, extra functional capabilities, which have led to large and complex vessels over the years. To a lesser extent, ship designers have put emphasis on the overall needs of customers. Securing a higher overall performance yield of the ship design solution than peer vessels out in the market is not a common practice among ship designers; they rather focus on a typical and traditional subset of performances. The implications and the consequences of such strategies in ship design have led to a growing need for a new set of design tools and project-making skills, a more extensive design process with different disciplines involved, and many iterations in the design development process.

Chemical identification of microfossils from the 1.88‐Ga Gunflint chert: Towards empirical biosignatures using laser ablation ionization mass spectrometer

AbstractIn this contribution, we investigated the chemical composition of Precambrian microfossils from the Gunflint chert (1.88 Ga) using a miniature laser ablation ionization mass spectrometer (LIMS) developed for in situ space applications. Spatially resolved mass spectrometric imaging (MSI) and depth profiling resulted in the acquisition of 68,500 mass spectra. Using single mass unit spectral decomposition and multivariate data analysis techniques, we identified the location of aggregations of microfossils and surrounding inorganic host mineral. Our results show that microfossils have unique chemical compositions that can be distinguished from the inorganic chert with high fidelity. Chemical depth profiling results also show that with LIMS microprobe data, it is possible to identify chemical differences between individual microfossils, thereby providing new insights about nature of early life. Analysis of LIMS spectra acquired from the individual microfossils reveals complex mineralization, which can reflect the metabolic diversity of the Gunflint microbiome. An intensity‐based machine learning model trained on LIMS Gunflint data might be applied for the future investigations of putative microfossils from silicified matrices, where morphological integrity of investigated structures is lost, and potentially in the investigation of rocks acquired from the Martian surface.

SERS-based DNA methylation profiling allows the differential diagnosis of malignant lymphadenopathy

This study highlights the potential of surface-enhanced Raman scattering (SERS) to differentiate between B-cell lymphoma (BCL), T-cell lymphoma (TCL), lymph node metastasis of melanoma (Met) and control (Ctr) samples based on the specific SERS signal of DNA extracted from lymph node tissue biopsy. Differences in the methylation profiles as well as the specific interaction of malignant and non-malignant DNA with the metal nanostructure are captured in specific variations of the band at 1005 cm−1, attributed to 5-methylcytosine and the band at 730 cm−1, attributed to adenine. Thus, using the area ratio of these two SERS marker bands as input for univariate classification, an area under the curve (AUC) of 0.70 was achieved in differentiating between malignant and non-malignant DNA. In addition, DNA from the BCL and TCL groups exhibited differences in the area of the SERS band at 730 cm−1, yielding an AUC of 0.84 in differentiating between these two lymphadenopathies. Lastly, using multivariate data analysis techniques, an overall accuracy of 94.7% was achieved in the differential diagnosis between the BCL, TCL, Met and Ctr groups. These results pave the way towards the implementation of SERS as a novel tool in the clinical setting for improving the diagnosis of malignant lymphadenopathy.

Surface-enhanced Raman spectral analysis for comparison of PCR products of hepatitis B and hepatitis C

BackgroundSurface-enhanced Raman spectroscopy is a reliable tool for identification and differentiation of two diseases showing similar symptoms, hepatitis B (HBV) and hepatitis C (HCV). ObjectivesTo develop a polymerase chain reaction technique (PCR) based SERS technique for differentiation of two human pathological conditions sharing the same symptoms using multivariate data analysis techniques e.g. principle component analysis (PCA) and partial least square discriminate analysis (PLS-DA). MethodsPCR products of HBV and HCV were differentiated by SERS using silver nanoparticles (AgNPs) as a SERS substrate. For this analysis, PCR products of both the diseases with predetermined viral loads were collected and analyzed under SERS instrument and unique SERS spectra of HBV and HCV was compared showing many differences at various points. Diseased classes of HBV and HCV and their negative control classes (viral load less than 1) were compared. PCR products of true healthy DNA and RNA were also compared, which were significantly separated. Moreover, SERS data was analyzed using multivariate data analysis techniques including principle component analysis (PCA) and partial least square discriminate analysis (PLS-DA) and differences were so prominent to observe. ResultsSERS spectral data of HBV and HCV showed clear differences and were significantly separated using PCA. Negative control samples of both disorders and their true healthy samples of DNA and RNA were separated according to 1st principle component. By analyzing data using partial least square discriminate analysis, differentiation of two disease classes was considered more valid with sensitivity, specificity and accuracy value of 96%, 94% and 98% respectively. Value of area under curve (AUROC) was 0.7527. ConclusionSERS can be employed for identification and comparison of two human pathological conditions sharing the same symptomology.

Extending TAM to Understand Library User Acceptance of E-Books in Tanzania

This study was conducted to examine antecedents that influence acceptance of e-books in Tanzania. The study extended TAM model with accessibility, perceived cost, and technological support to develop a conceptual model that was used to examine acceptance of e-books in Tanzania. Quantitative study was conducted using self-administered questionnaire in which snowball technique was used to identify potential respondents. The questionnaires were distributed using online Google form, and 314 usable responses were collected. A multivariate data analysis technique known as structural equation modeling (SEM) was employed to establish relationships between the hypothesized relationships. The findings show that perceived ease-of-use, accessibility, and technical support have significant effect on acceptance of e-books in Tanzania. Several implications that will lead to acceptance of e-books in Tanzania are provided.

Surface-enhanced Raman spectroscopy for comparison of serum samples of typhoid and tuberculosis patients of different stages

BackgroundSurface-enhanced Raman spectroscopy (SERS) is a reliable tool for the identification and differentiation of two different human pathological conditions sharing the same symptomology, typhoid and tuberculosis (TB). ObjectivesTo explore the potential of surface-enhanced Raman spectroscopy for differentiation of two different diseases showing the same symptoms and analysis by principal component analysis (PCA) and partial least square discriminate analysis (PLS-DA). MethodsSerum samples of clinically diagnosed typhoid and tuberculosis infected individuals were analyzed and differentiated by SERS using silver nanoparticles (Ag NPs) as a SERS substrate. For this purpose, the collected serum samples were analyzed under the SERS instrument and unique SERS spectra of typhoid and tuberculosis were compared showing notable spectral differences in protein, lipid and carbohydrates features. Different stages of the diseased class of typhoid (Early acute and late acute stage) and tuberculosis (Pulmonary and extra-pulmonary stage) were compared with each other and with healthy human serum samples, which were significantly separated. Moreover, SERS data was analyzed using multivariate data analysis techniques including principal component analysis (PCA) and partial least square discriminate analysis (PLS-DA) and differences were so prominent to observe. ResultsSERS Spectral data of typhoid and tuberculosis showed clear differences and were significantly separated using PCA. SERS spectral data of both stages of typhoid and tuberculosis were separated according to 1st principle component. Moreover, by analyzing data using partial least square discriminate analysis, differentiation of two disease classes were considered more valid with a 100% value of sensitivity, specificity and accuracy. ConclusionSERS can be employed for identification and comparison of two different human pathological conditions sharing same symptomology.

Financial Inclusion and Quality of Life: Empirical Evidences from Indian States with Special Reference to West Bengal

The conceptual connection between financial inclusion and quality of life (QOL) can be realized by a two-way relationships. On the one hand, financial inclusion induces QOL, while an improvement in QOL facilitates in generating demand for financial services, on the other hand. Even though several studies seek to find out the role of finance in the well-being of the population (especially human development), this article concentrates on QOL to eliminate the financial attributes of development (as captured by income dimension in Human Development Index). In this sense, this study addresses the research gap in the existing literature by establishing the relationship between financial inclusion and QOL. Specifically, the article attempts to explain the two-way tie-up between the financial inclusion and the QOL in India in the context of Indian states, in general, and West Bengal, in particular. Canonical correlation (CC; a multivariate data analysis technique) is used to estimate the relation between the financial inclusion and QOL. Empirical results suggest that western and the southern Indian states excel in the attainment of education, health and other amenities-based indicators of QOL. The conditions of the eastern part of the country in case of financial inclusion and the QOL are not at all satisfactory. In case of West Bengal, Kolkata being the state capital is performing well in both the factors. CC results suggest a significant association between the financial inclusion and QOL across Indian states. The deposit account of financial inclusion indicator and the infant survival rate of QOL indicator are playing a pivotal role in the relationship (both the Indian states and districts of West Bengal as well). This article establishes the effectiveness of the demand following approach of financial inclusion than that of supply leading approach. As the demand-side aspect of financial inclusion is becoming more important to the policymakers, the next policy priority of financial inclusion measures could be the generation of awareness on the financial services through financial literacy. JEL Codes: G2, O15, C39

Characterization and prediction of viral loads of Hepatitis B serum samples by using surface-enhanced Raman spectroscopy (SERS)

BackgroundRaman spectroscopy is a promising technique to analyze the body fluids for the purpose of non-invasive disease diagnosis. ObjectivesTo develop a surface-enhanced Raman spectroscopy (SERS) based method for qualitative and quantitative analysis of hepatitis B viral (HBV) infection from blood serum samples. MethodsClinically diagnosed hepatitis B virus (HBV) infected serum samples of patients of different levels of viral loads have been subjected for SERS analysis in comparison with the healthy ones by using silver nanoparticles (Ag NPs) based SERS substrates. The SERS measurements were performed on blood serum samples of 11 healthy and 32 clinically diagnosed HBV patients of different viral load levels of different exponentials including (101, 102 called as low level), (103, 104 called as medium level) and (105, 108 called as high level). Furthermore, multivariate data analysis techniques, Principal Component Analysis (PCA) and Partial Least Square Regression (PLSR) were also performed on SERS spectral data. ResultsThe SERS spectral features due to biochemical changes in HBV positive serum samples associated with the increasing viral loads were established which could be employed for HBV diagnostic purpose. PCA was found helpful for the differentiation between SERS spectral data of serum samples of different levels of HBV infection and healthy individuals. PLSR model developed with standard samples of known viral loads for predicting the viral loads of blind/unknown samples with 99% predicted accuracy. ConclusionSERS can be employed for qualitative and quantitative analysis of HBV infection from blood serum samples.

Surface-enhanced Raman spectroscopy for identification of food processing bacteria

Food processing bacteria play important role in providing flavors, ingredients and other beneficial characteristics to the food but at the same time some bacteria are responsible for food spoilage. Therefore, quick and reliable identification of these food processing bacteria is very necessary for the differentiation between different species which may help in the development of more useful food processing methodologies. In this study, analysis of different bacterial species (Lactobacillus fermentum, Fructobacillus fructosus, Pediococcus pentosaceus and Halalkalicoccus jeotgali) was performed with our in-house developed Ag NPs-based surface-enhanced Raman spectroscopy (SERS) method. The SERS spectral data was analyzed by multivariate data analysis techniques including principal component analysis (PCA) and partial least square discriminant analysis (PLS-DA). Bacterial species were differentiated on the basis of SERS spectral features and potential of SERS was compared with the Raman spectroscopy (RS). SERS along with PCA and PLS-DA was found to be an efficient technique for identification and differentiation of food processing bacterial species. Differentiation with accuracy of 99.5% and sensitivity of 99.7% was depicted by PLS-DA model using leave one out cross validation.

Brand Positioning and Segmentation of Sneakers through Multi-Dimensional Customer Experience Analysis

A huge corpus of valuable information on customer experience is available as unstructured form in customer reviews on e-commerce websites. Multivariate data analysis techniques are effective in uncovering hidden patterns and segments in structured data. A major challenge is to convert the unstructured data into a structured form for applying multivariate techniques. In this article, we have provided a text analysis based approach coupled with multivariate techniques to uncover the sentiment of various features associated with different brands and to determine the brand positions and segments through perceptual mapping and cluster analysis.

Rapid and sensitive discrimination among carbapenem resistant and susceptible E. coli strains using Surface Enhanced Raman Spectroscopy combined with chemometric tools

BackgroundRaman spectroscopy is a powerful technique for the robust, reliable and rapid detection and discrimination of bacteria. ObjectivesTo develop a rapid and sensitive technique based on surface-enhanced Raman spectroscopy (SERS) with multivariate data analysis tools for discrimination among carbapenem resistant and susceptible E. coli strains. MethodsSERS was employed to differentiate different strains of carbapenem resistant and susceptible E. coli by using silver nanoparticles (Ag NPs) as a SERS substrate. For this purpose, four strains of carbapenem resistant and three strains of carbapenem susceptible E. coli were analyzed by comparing their SERS spectral signatures. Furthermore, multivariate data analysis techniques including Principal Component Analysis (PCA), Hierarchical Cluster Analysis (HCA) and Partial Least Squares-Discriminant Analysis (PLS-DA) were performed over the spectral range of 400–1800 cm−1 (fingerprint region) for the identification and differentiation of different E. coli strains. ResultsThe SERS spectral features associated with resistant development against carbapenem antibiotics were separated by comparing each spectrum of susceptible strains with each resistant strain. PCA and HCA were found effective for the qualitative differentiation of all the strains analysed. PLS-DA successfully discriminated the carbapenem resistant and susceptible E. coli pellets on the strain level with 99.8 % sensitivity, 100 % specificity, 100 % accuracy and 86 % area under receiver operating characteristic (AUROC) curve. ConclusionSERS can be employed for the rapid discrimination among carbapenem resistant and susceptible strains of E. coil.

Bundles of Lean Automation practices and principles and their impact on operational performance

The objective of this paper is two-fold. First, it aims at empirically validating bundles of Lean Automation (LA) practices and principles. Second, it verifies their impact on operational performance improvement. For that, we surveyed 110 manufacturers who have been concomitantly implementing Lean Production (LP) and Industry 4.0 (I4.0). The collected responses were examined using multivariate data analysis techniques. We identified three bundles of LA named according to the focus of the corresponding practices (i.e. workplace, value stream and extended value stream), and two bundles of LA principles (sociocultural- and technology-oriented). Our findings indicated that LA practices positively mediate the relationship between LA principles and the company's operational performance, especially when considering practices applied at a workplace or value stream level. As the integration of I4.0 into LP becomes widespread, the LA paradigm tends to be a more relevant topic in the operations management field. Hence, more profound comprehension of LA practices and principles, and their implications on companies' operations may help achieve higher performance, setting the proper managerial expectations.

Business Dynamics in Recovery Times: A Comparative Perspective on Manufacturing Firms’ Performance in the European Union

Abstract Our paper investigates the gaps in performance in the manufacturing sector between Western and Eastern European countries and attempts to analyze how enterprises from these two parts of Europe have tackled recovery after the Global financial crisis of 2007-2009. We uncover the patterns of performance in the after-crisis period and offer insights into the prospects of the manufacturing sector in the European Union, faced nowadays with a new recovery, after the coronavirus crisis. Moreover, we study these patterns in industries with different technological levels. We have selected five performance variables, namely Turnover growth rate, Turnover per employee, Wage-adjusted labor productivity, Gross operating rate, and Investment rate, and employed statistical cluster analysis, which is a multivariate data analysis technique that can detect these patterns in performance, in both its approaches: hierarchical and k-means clustering. Our findings show that the almost perfect groupings of businesses from Western, more developed economies, and Eastern, less developed ones, in all industries, with the notable exception of Portugal, are rather striking, regardless of the technological level of industries. We show that Eastern EU businesses were not the worst performers in the after-crisis period, but rather on the contrary. Certainly, they are smaller in size but have enjoyed higher labor productivity and profitability, as well as higher investment rates in all industries. This points towards a higher dynamism of smaller-sized businesses in general, and Eastern EU located ones, in particular, in the years after the Global financial crisis, which has been reflected in superior performance.

Determinants of Continuance Intention to Use Mobile Wallet Services : Light Users vs Heavy Users

Mobile technologies and rapid popularity & usage among the businesses and consumers have led to revolutionary changes in the way the goods and services are marketed and consumed. The impact of mobile based communication strategies among modern marketers are more commonly witnessed with the revolutionary growth of mobile applications and their usage. One of the significant developments of mobile technologies that prove to be beneficial to both businesses and customers is the introduction of mobile payment systems, otherwise known as mobile wallet applications. Despite many mobile wallet users in India, it is highly challenging for service providers to understand the motivations behind customers to be loyal because of divergent customer segments and their needs. Hence, this study aimed at identifying the factors that determine the mobile wallet users’ intentions to continue using a specific service provider and how this behaviour would be different across the customer groups based on their usage. This paper proposed a theoretical research framework and validated the same with empirical data with the help of multivariate data analysis techniques, including exploratory factor analysis, t-test, and MANOVA. Perceived usefulness, perceived ease of use, perceived enjoyment, and satisfaction were found to be the significant factors influencing users' intention to continue using mobile wallet applications. Finally, the results revealed that light users and heavy users differed significantly in terms of their overall post-adoption behaviour on mobile wallet applications.

The Effect of Entrepreneurial Orientation on Business Performance of Small and Medium Scale Enterprises: A Study of Amhara Region of Ethiopia

Entrepreneurial Orientation – Performance linkage researches have gained the attention of researchers for the last some decades. Many of previous studies show that Entrepreneurial Orientation (EO) as a strategic orientation is a key element for firm success, but few studies have reported an insignificant relationship. This study investigated the effect of overall entrepreneurial orientation on the performance of Small and Medium Enterprises (SMEs) in Amhara Region, Ethiopia. Data was collected through structured questionnaire from owners/managers of SMEs. A multivariate data analysis technique of Structural Equation Modelling (SEM) was employed to analyze the data. The finding indicates that overall EO significantly and positively affects the business performance of small and medium enterprises. The findings are expected to benefit for owner/managers of SMEs to be more entrepreneurial oriented to improve their business performance and in order to survive in the intensive competitive market environment. For governmental and nongovernmental organizations, it helps to design strategy to make SMEs more entrepreneurial. This study also provides an opportunity to expand the research on other sectors.

Method to appraise financial and productivity indicators in small and medium enterprises in the tourist lodging sector of Cartagena-Colombia applying discriminant multivariate analysis

This research article presents the results of the appraising of the financial and productivity indicators of the small and medium enterprises of the tourist lodging sector in the City of Cartagena-Colombia. The multivariate discriminant data analysis (MDA) technique was used to determine the significance and classification of the groups of selected variables within the group of 75 SMEs that reported their financial statements at the Cartagena Chamber of Commerce for the years 2015 and 2018. The findings found contrast with the sustained good performance for the analysis period of the financial indicators, gross margin (IF1), operating margin (IF2) and net margin (IF3) and productivity, operating profit ratio and added value (IP2), added value ratio and working capital (IP4) and operating profit ratio and working capital (IP5). This is evidenced by the marking out and contribution of the tourist lodging sector in the business economic dynamics of the City of Cartagena.

An integrated principal component and reduced multivariate data analysis technique for detecting DDoS attacks in big data federated clouds

An integrated principal component and reduced multivariate data analysis technique for detecting DDoS attacks in big data federated clouds