Multivariate Principal Component Research Articles

Design of ensemble recurrent model with stacked fuzzy ARTMAP for breast cancer detection

PurposeIn time and accurate detection of cancer can save the life of the person affected. According to the World Health Organization (WHO), breast cancer occupies the most frequent incidence among all the cancers whereas breast cancer takes fifth place in the case of mortality numbers. Out of many image processing techniques, certain works have focused on convolutional neural networks (CNNs) for processing these images. However, deep learning models are to be explored well.Design/methodology/approachIn this work, multivariate statistics-based kernel principal component analysis (KPCA) is used for essential features. KPCA is simultaneously helpful for denoising the data. These features are processed through a heterogeneous ensemble model that consists of three base models. The base models comprise recurrent neural network (RNN), long short-term memory (LSTM) and gated recurrent unit (GRU). The outcomes of these base learners are fed to fuzzy adaptive resonance theory mapping (ARTMAP) model for decision making as the nodes are added to the F_2ˆa layer if the winning criteria are fulfilled that makes the ARTMAP model more robust.FindingsThe proposed model is verified using breast histopathology image dataset publicly available at Kaggle. The model provides 99.36% training accuracy and 98.72% validation accuracy. The proposed model utilizes data processing in all aspects, i.e. image denoising to reduce the data redundancy, training by ensemble learning to provide higher results than that of single models. The final classification by a fuzzy ARTMAP model that controls the number of nodes depending upon the performance makes robust accurate classification.Research limitations/implicationsResearch in the field of medical applications is an ongoing method. More advanced algorithms are being developed for better classification. Still, the scope is there to design the models in terms of better performance, practicability and cost efficiency in the future. Also, the ensemble models may be chosen with different combinations and characteristics. Only signal instead of images may be verified for this proposed model. Experimental analysis shows the improved performance of the proposed model. This method needs to be verified using practical models. Also, the practical implementation will be carried out for its real-time performance and cost efficiency.Originality/valueThe proposed model is utilized for denoising and to reduce the data redundancy so that the feature selection is done using KPCA. Training and classification are performed using heterogeneous ensemble model designed using RNN, LSTM and GRU as base classifiers to provide higher results than that of single models. Use of adaptive fuzzy mapping model makes the final classification accurate. The effectiveness of combining these methods to a single model is analyzed in this work.

Evaluation of the physiological potential of Panicum maximum seeds by multivariate analysis

The aim of this work was to identify efficient vigor tests for differentiating the seed lots, forecasting seedling emergence in the field and assessing the physiological quality of Panicum maximum seeds. 12 seed lots from the cultivar Tanzania and 11 seed lots from the cultivar Massai were evaluated for water content, germination, first count and germination speed index, emergence and first emergence count of seedlings in sand, root length and shoot length, analysis of SVIS® images (seedling length, vigor and uniformity index) and seedling emergence in the field. The work was conducted in a completely randomized design for tests performed in the laboratory and in randomized blocks for tests in the field. The data were subjected to analysis of variance and the means compared by Scott Knott's test at 5% probability and statistical multivariate clustering analysis and principal components analysis. The shoot and root length tests are efficient for the evaluation of the physiological potential of P. maximum cv. Massai, while the seedling length, vigor index and growth uniformity index tests using image analysis, seedling emergence in sand and first seedling emergence count in sand are efficient in assessing the physiological potential of seeds of P. maximum cv. Tanzania, and providing information similar to that of seedling emergence in the field.

3D imaging of single bacterial cells using surface-enhanced Raman spectroscopy with a multivariate curve resolution model

Imaging biomolecules within a single bacterial cell is crucial for understanding cellular genetic mechanisms. Herein, we exploited a surface-enhanced Raman spectroscopy (SERS) imaging strategy for single cell analysis. Cellular biosynthesized Ag nanoparticles (NPs) provided the necessary enhancement for SERS imaging. Multiple complementary techniques, including high-resolution transmission electron microscopy (HR-TEM), high-angle annular dark-field (HAADF)-scanning transmission electron microscopy (STEM), and energy-dispersive X-ray spectroscopy (EDX), were used to characterize the biogenic Ag NPs in cells. Three-dimensional SERS imaging maps displayed spectral information of biomolecules within the single cell. The multivariate curve resolution (MCR) model and principal component analysis (PCA) model were used to analyze the cellular SERS imaging maps. The MCR model, with a specific constraint of non-negativity, resulted in meaningful identification of biomolecules associated with Ag reduction. Focusing on the molecular level reveals that Pantoea sp. IMH utilizes several mechanisms to synthesize Ag NPs, including cytoplasm reduction by glucose or nicotinamide adenine dinucleotide (NADH)-dependent reductase, and extracellular reduction by an electron transfer chain containing quinone and cytochrome C. Our results shed new light on the Ag NP biosynthesis mechanism and single cell Raman analysis.

Multidimensional economic indicators and multivariate functional principal component analysis (MFPCA) in a comparative study of countries’ competitiveness

Multidimensional economic indicators and multivariate functional principal component analysis (MFPCA) in a comparative study of countries’ competitiveness

Functional ANOVA approaches for detecting changes in air pollution during the COVID-19 pandemic

Faced with novel coronavirus outbreak, the most hard-hit countries adopted a lockdown strategy to contrast the spread of virus. Many studies have already documented that the COVID-19 control actions have resulted in improved air quality locally and around the world. Following these lines of research, we focus on air quality changes in the urban territory of Chieti-Pescara (Central Italy), identified as an area of criticality in terms of air pollution. Concentrations of hbox {NO}_{{2}}, hbox {PM}_{{10}}, hbox {PM}_{2.5} and benzene are used to evaluate air pollution changes in this Region. Data were measured by several monitoring stations over two specific periods: from 1st February to 10 th March 2020 (before lockdown period) and from 11st March 2020 to 18 th April 2020 (during lockdown period). The impact of lockdown on air quality is assessed through functional data analysis. Our work makes an important contribution to the analysis of variance for functional data (FANOVA). Specifically, a novel approach based on multivariate functional principal component analysis is introduced to tackle the multivariate FANOVA problem for independent measures, which is reduced to test multivariate homogeneity on the vectors of the most explicative principal components scores. Results of the present study suggest that the level of each pollutant changed during the confinement. Additionally, the differences in the mean functions of all pollutants according to the location and type of monitoring stations (background vs traffic), are ascribable to the hbox {PM}_{{10}} and benzene concentrations for pre-lockdown and during-lockdown tenure, respectively. FANOVA has proven to be beneficial to monitoring the evolution of air quality in both periods of time. This can help environmental protection agencies in drawing a more holistic picture of air quality status in the area of interest.

Potential of amyl alcohol mixtures derived from Scenedesmus quadricauda microalgae biomass as third generation bioenergy for compression ignition engine applications using multivariate-desirability analysis

ABSTRACT This study is aimed at synthesizing amyl alcohol from Scenedesmus quadricauda third-generation microalgae biomass which is cultivated sustainably using reactor combinations and Ehrlich biosynthetic pathway. Systematic experimentation was carried out on a compression ignition engine at maximum throttle pedal position condition to assess the potential of amyl alcohol-diesel mixture combinations and to assess its environmental impact at different loads. Optimum mixture proportion of amyl alcohol-diesel blends were identified statistically using multivariate principal component and desirability analysis based on the experimental data acquired from the engine testbed. From the study, it was inferred that a dry biomass yield rate of 4.2 g/L was achieved when the light intensity was maintained at ~51-56 μmol/m2/s in the in-situ raceway pond while cultivating Scenedesmus quadricauda. Experimentation showed that the fuel consumption reduced by 72.4% for AA20D80 at low-high load conditions. Improved combustion characteristics are achieved at higher concentrations of amyl alcohol operating at medium-high loads. The lowest smoke and NOX emission of 52.85 mg/m3 and 104 ppm was observed for AA40D60 at no-load condition. Statistical analysis revealed that the optimal amyl alcohol-diesel mixture combinations to be AA20D80 with combined desirability of 0.497.

Assessing the Validity of Mojave Desert Lake Mohave and Silver Lake Projectile-Point Types

ABSTRACT This study uses univariate and multivariate statistical analyses to assess whether Lake Mohave and Silver Lake projectile points from the Mojave Desert of western North America are statistically discrete, recognizable, and valid types. Univariate analyses run on 400 Lake Mohave and Silver Lake points reveal significant differences in six linear and three angle measures, though overlapping measurements pose difficulties for differentiating the types. To address the problem of overlapping measurements, multivariate principal components and discriminant function analyses were employed to statistically define groups of related attributes. Size- and shape-based attributes of the stem distinguish approximately 80% of the sample to type and reveal region-wide consistency among archaeologists who have classified these points to type. Though this study did not entirely overcome the challenge of using metrics to differentiate Lake Mohave from Silver Lake points, key discriminating variables are now better defined and the types deemed valid.

In situ X-ray absorption study of Cu species in Cu-CHA catalysts for NH3-SCR during temperature-programmed reduction in NO/NH3

Ammonia-mediated selective catalytic reduction (NH3-SCR) using Cu-exchanged chabazite zeolites as catalysts is one of the leading technologies for NOx removal from exhaust gases, with CuII/CuI redox cycles being the basis of the catalytic reaction. The amount of CuII ions reduced by NO/NH3 can be quantified by the consumption of NO during temperature-programmed reduction experiments (NO-TPR). In this article, we show the capabilities of in situ X-ray absorption near-edge spectroscopy (XANES), coupled with multivariate curve resolution (MCR) and principal component analysis (PCA) methods, in following CuII/CuI speciation during reduction in NO/NH3 after oxidation in NO/O2 at 50 °C on samples with different copper loading and pretreatment conditions. Our XANES results show that during the NO/NH3 ramp CuII ions are fully reduced to CuI in the 50–290 °C range. The number of species involved in the process, their XANES spectra and their concentration profiles as a function of the temperature were obtained by MCR and PCA. Mixed ligand ammonia solvated complexes [CuII(NH3)3(X)]+ (X = OH−/O− or NO3−) are present at the beginning of the experiment, and are transformed into mobile [CuI(NH3)2]+ complexes: these complexes lose an NH3 ligand and become framework-coordinated above 200 °C. In the process, multiple CuII/CuI reduction events are observed: the first one around 130 °C is identified with the reduction of [CuII(NH3)3(OH/O)]+ moieties, while the second one occurs around 220–240 °C and is associated with the reduction of the ammonia-solvated Cu-NO3− species. The nitrate concentration in the catalysts is found to be dependent on the zeolite Cu loading and on the applied pretreatment conditions. Ammonia solvation increases the number of CuII sites available for the formation of nitrates, as confirmed by infrared spectroscopy.

Novel Biomarkers to Distinguish between Type 3c and Type 2 Diabetes Mellitus by Untargeted Metabolomics.

Pancreatogenic diabetes mellitus (T3cDM) is a highly frequent complication of pancreatic disease, especially chronic pancreatitis, and it is often misdiagnosed as type 2 diabetes mellitus (T2DM). A correct diagnosis allows the appropriate treatment of these patients, improving their quality of life, and various technologies have been employed over recent years to search for specific biomarkers of each disease. The main aim of this metabolomic project was to find differential metabolites between T3cDM and T2DM. Reverse-phase liquid chromatography coupled to high-resolution mass spectrometry was performed in serum samples from patients with T3cDM and T2DM. Multivariate Principal Component and Partial Least Squares-Discriminant analyses were employed to evaluate between-group variations. Univariate and multivariate analyses were used to identify potential candidates and the area under the receiver-operating characteristic (ROC) curve was calculated to evaluate their diagnostic value. A panel of five differential metabolites obtained an area under the ROC curve of 0.946. In this study, we demonstrate the usefulness of untargeted metabolomics for the differential diagnosis between T3cDM and T2DM and propose a panel of five metabolites that appear altered in the comparison between patients with these diseases.

Características morfológicas de la Cabra Criolla del Sur de Mendoza, Argentina

The objective of this work was to determine the values of zoometric variables that define the Creole Goat of South Mendoza (Mendoza province, Argentina). Samples were taken at commercial goat farms located in South and North regions and in seven locations of San Rafael and Malargue departments of Mendoza. Eight hundred and thirty five Creole goats were sampled, with ages from two teeth up to tooth cutter, using the method proposed by FAO (2000). The variables studied were Head width (HW), Head length (HL), Height to the end of Neck (HN), Thoracic perimeter (TP), Body length (BL), Rump width (RW); Height to the Rump (HR) and Shank Perimeter (SP). The variables were analyzed using multivariate techniques principal components, discriminants, canonical populations) and clusters based on Euclídeus distances and their graphic representation as dendrograms. The results showed that all variables followed a normal distribution. In turn, uniformity in the variables was found, since Coefficient of Variation was lower than 9% in all cases. The variables that best describe the model were HN, TP, RW, SP and HN, which represent 50% of the total data variability. The dendogram showed two mixed groups where variables of locations of North and South regions were found, finding the smallest eucledian distance among locations of the South region. The Figure of the Analysis of Principal Components showed that dispersion of the variables behaves in an homogeneous way, without generating groups or subpopulations. Variable homogeneity found in this work would allow to assume that the Creole Goats of the South Mendoza would belong to a common genetic stem that would characterize them as members of a defined racial group.

Assessing environmental impacts in using waste steel slags as construction materials in a highly industrialized area

Steel slags are side products of iron and steel industry and are suitable for a variety of applications. In the present paper, the assessment of the hazard associated with steel slags used as harbor piers backfilling material was performed. The analyses were conducted on 64 samples (soil and groundwater) collected in 17 sites located in one of the piers of Taranto city (South Italy), a highly industrialized area, where steel slags have been used in the past as backfill material. The obtained data were compared to the national threshold levels and International Indexes (Igeo) for assessing anthropogenic impacts. The first comparative analysis revealed that only vanadium in soil and aluminum in groundwater, associated with high pH values in both matrixes (soil and groundwater) resulted of high concern. Nevertheless, Igeo values and the coupling of comparative analysis with multivariate techniques Principal component analysis and Hierarchical Cluster Analyses revealed that also Sb, Pb, Cr and Zn concentration have an extremely high hazard degree. All analyses confirm the impact of steel slags on contaminant leaching processes due to the strong alkaline nature of their leachates and the presence of strong relationship between metal mobility and other physical-chemical parameters related to the peculiar environmental context. These findings suggest that the use of such materials, albeit economically recommended, should be adopted under severe monitoring of all parameters that could promote the mobility of metals and other contaminants.

Multi-sensor prognostics modeling for applications with highly incomplete signals

Multi-stream degradation signals have been widely used to predict the residual useful lifetime of partially degraded systems. To achieve this goal, most of the existing prognostics models assume that degradation signals are complete, i.e., they are observed continuously and frequently at regular time grids. In reality, however, degradation signals are often (highly) incomplete, i.e., containing missing and corrupt observations. Such signal incompleteness poses a significant challenge for the parameter estimation of prognostics models. To address this challenge, this article proposes a prognostics methodology that is capable of using highly incomplete multi-stream degradation signals to predict the residual useful lifetime of partially degraded systems. The method first employs multivariate functional principal components analysis to fuse multi-stream signals. Next, the fused features are regressed against time-to-failure using (log)-location-scale regression. To estimate the fused features using incomplete multi-stream degradation signals, we develop two computationally efficient algorithms: subspace detection and signal recovery. The performance of the proposed prognostics methodology is evaluated using simulated datasets and a degradation dataset of aircraft turbofan engines from the NASA repository.

A methodology for apportioning federal SUS resources: the health needs index.

ABSTRACTOBJECTIVE To present a methodology for apportioning Union resources to the federative units (FU – 26 states and one federal district) within the Brazilian Unified Health System (SUS) based on health needs measured by demographic, socioeconomic, epidemiological and geographical dimensions.METHODS The apportionment methodology proposal prioritizes the health needs axis, based on Law 141/2012. We adopted a proxy of needs that measures relative inequalities between, socioeconomic, geographic demographic and epidemiological conditions of the populations of the Brazilian Federative Units (FU) for 2015. We first used an adjustment so that the populations of the 27 FU are corrected by their relative needs regarding age and gender. To calculate the health needs axis, the multivariate techniques factorial analysis and principal components were used, and, based on such correction, we applied the health needs index. Subsequently, this index was applied to simulate the resources that should be transferred by the Ministry of Health to states in 2015.RESULTS As we made the methodological choice of transferring a single per capita amount to all states, so the proposal required population correction. Thus, in the analysis of health needs, the FUs that had their population corrected by a factor higher than the national average because of their greater relative need, were the states of: Maranhão, Piauí, Alagoas, Paraíba, Ceará, Pará, Bahia, Acre, Pernambuco, Rio Grande do Norte, Sergipe, Amazonas, Tocantins and Roraima. For the simulation aggregating all the financing blocks, without reducing the resources already distributed to the remaining states in 2015, indicated the additional need of R$ 4.6 billion.CONCLUSIONS The proposal addresses the absence of studies presenting quantitative simulations of federal resources distribution within the scope of SUS to the FUs, based on the apportionment criteria defined by Law 141/2012, in order to contribute to the reduction health inequalities and mitigate the effects of the economic crisis.

Morpho-agronomical characterisation of local and international sweetpotato germplasm from the South African collection

Considerable genetic diversity is present amongst the sweet potato (Ipomoea batatas (L.) Lam.) accessions that can be selected for breeding. The Agricultural Research Council of South Africa maintains 360 sweetpotato accessions that were bred locally, and introductions from various regions of the world sourced for breeding purposes. The objective of this study was to determine the phenotypic diversity of 73 accessions, using morpho-agronomical characteristics to complement data on protein content of the accessions towards breeding for improved nutritional content. A field study was conducted using a randomized complete block design. Quantitative and qualitative data were collected using 38 of the international standardized descriptors for sweetpotato. Agronomic traits including root yield and biomass, and quality traits such as dry matter, β-carotene and protein contents, and root oxidation were determined. Multivariate clustering analysis and principal component analysis grouped the 73 sweetpotato accessions into three main groups. The major traits clustering the accessions were leaf lobe type, central leaf lobe shape, vine tip pubescence, vine pigmentation colour, and flowering ability. Of the 73 accessions 16.4% produced few flowers, 21.9% had intermediate flowering and 9.5% flowered abundantly. This study provided baseline information on the phenotypic diversity of South African sweetpotato collections for effective utilization and future breeding.

Soil moisture estimation using Bayesian Maximum Entropy algorithm from FY3-B, MODIS and ASTER GDEM remote-sensing data in a maize region of HeBei province, China

ABSTRACT The Bayesian Maximum Entropy (BME) algorithm that can model larger-scale spatial heterogeneity and integrate multiple types of data is a better spatial estimation algorithm in avoiding the circular problem and improving the estimation accuracy of parameters than original empirical geostatistic and spatial statistic methods. To obtain higher-resolution and higher-accuracy soil moisture (SM) data, this study used the BME algorithm to integrate the multiple environmental factor variables related to SM as auxiliary data (11 data sets in total) such as the Normalized Difference Vegetation Index (NDVI), Land Surface Temperature (LST), and Albedo products with 1 kilometre (km) grid resolution from the Moderate-resolution Imaging Spectroradiometer (MODIS), the Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Model (ASTER GDEM) satellite data, and the Slope, Aspect, Plan curvature, Profile curvature, Surface roughness, Wetness index, Relief amplitude products generated from ASTER GDEM data. First, the FengYun 3-B satellite (FY3-B) SM product with 25 km grid resolution was downscaled to 1 km grid resolution using the visible, infrared (IR), and microwave fusion method. The downscaled FY3-B SM data and auxiliary data were used to generate the weighted probability soft data (SD) under four cases, case 1: 500 sample points of SD, case 2: 450 sample points of SD, case 3: 400 sample points of SD, and case 4: 350 sample points of SD. Particularly, we used two methods, the multivariable correlation analysis method and principal component analysis (PCA) method to get the weight values of environmental factor variables, i.e. NDVI, LST, Albedo, the Digital Elevation Model (DEM), Slope, Aspect, Plan curvature, Profile curvature, Surface roughness, Wetness index, and Relief amplitude with SM. Then, the in-situ SM measurements as hard data (HD) were used to calibrate the weighted probability SD in the procedure of BME SM estimation. The SM under four cases were estimated from the weighted probability SD and HD using BME algorithm. Finally, comparisons of the estimated SM using BME algorithm with in-situ SM measurements at maize study area were carried out in this study. Our results indicated that the accuracy of the estimated SM using BME algorithm, the root-mean-square error (RMSE) = 0.049 cm3 cm−3, the correlation coefficient (r) = 0.639, the unbiased RMSE (RMSEu) = 0.047 cm3 cm−3, and Bias = 0.002 cm3 cm−3, under case 1 based on PCA method was obviously better than the downscaled FY3-B SM (RMSE = 0.079 cm3 cm−3, r = 0.096, RMSEu = 0.069 cm3 cm−3, and Bias = 0.039 cm3 cm−3) that was generated by the visible, IR, and microwave fusion method. We concluded that integrating auxiliary data into SM estimation using BME algorithm could further improve the downscaled FY3-B SM that was generated by visible, IR, and microwave fusion method.

Geo-multivariate analysis of crop yield and soil properties of students industrial training farm in Ado Ekiti, Southwest Nigeria

ABSTRACT Relating crop yield to soil properties at the field scale is a tool in site-specific management planning for sustainable crop production. This study investigated the relationship between some soil properties and the yield of three crops (cowpea, maize and cassava + maize) grown on Students Industrial Work Experience Scheme Training Farm in Ado Ekiti, Nigeria using descriptive, multivariate and geostatistical techniques. Soil samples were collected from 0 - 20 cm depth at the centre of 184 grids from a 3.0 ha field to determine some soil physico-chemical properties. Crop yield was collected from demarcated areas in each grid. The descriptive statistics indicated that the soil properties and yield varied widely under the different crops. The cross-variograms showed moderate to strong spatial dependence for the soil properties. The multivariate principal component and hierarchal cluster analyses showed sand, clay, organic matter, bulk density, pH, electrical conductivity, and available phosphorus as the minimum dataset for upscaling soil quality assessment program. The multivariate and geostatistical analyses revealed significant relationships between soil properties and crop yield. These findings will enable optimization of the farmer’s field through specific soil and nutrient management, reduce crop yield variability, and promote a safe environment.

A Reinforcement Learning Method Using Multifunctional Principal Component Analysis for Human-Like Grasping

Postural synergies allow a rich set of hand configurations to be represented in lower dimension space compared to the original joint space. In our previous works, we have shown that this can be extended to trajectories thanks to the multivariate functional principal component analysis, obtaining a set of basis functions able to represent grasping movements learned from human demonstration. In this article, we introduce a human cognition-inspired approach for generalizing and improving robot grasping skills in this motion synergies subspace. The use of a reinforcement learning (RL) algorithm allows the robot to explore the surrounding space and improve its capability to reach and grasp objects. The learning method is the policy improvement with path integrals, running in the policy space. Bootstrapped with synergy coefficients obtained from neural networks, the policy reward is based on a force closure grasp quality index computed at the end of the task, measuring how a firm is the grip. We finally show that combining neural networks and RL allows the robot manipulator to have a good initial estimate of the grasping configuration and faster convergence to an optimal grasp with respect to a database approach, the latter a less general solution in presence of new objects.

Lipophilicity of tryptophan, its metabolites and derivatives measured by thin-layer chromatography

Fourteen compounds, including tryptophan and its metabolites: kynurenine pathway products (kynurenine, kynurenic acid, 3-hydroxykynurenine), auxin pathway products (indoleacetic acid, indolepropionic acid, indolebutyric acid, indolepyruvic acid), serotonin pathway products (5-hydroxytryptophan, serotonin, melatonin), tryptamine, as well as two synthetic auxin analogs (naphtyl-1-acetamide and naphthalene-1-acetic acid) were chromatographed on RP8 plates using binary mobile phases containing 8 concentrations of methanol as a modifier. The resulting retention data was fitted to linear Soczewiński–Wachtmeister models, obtaining experimental lipophilicity indices, which were then compared with computational lipophilicities calculated with 6 algorithms (ALOGPs, AClogP, ALOGP, MLOGP, XLOGP2 and XLOGP3) using regression and multivariate (principal component analysis) approach. It was confirmed that hydroxylation lowers the lipophilicity, whereas adding one homologous aliphatic CH2, exchange of amino group to keto group (or its loss) increases the lipophilicity. Kynurenic acid has lower lipophilicity than kynurenine under investigated conditions due to its ionization, which is opposite to computational results based on non-ionized molecule. The obtained results can be useful in explaining the biochemical and pharmacological phenomena related to these compounds.

Board of directors, capital structure, investment decisions and firm-performance: An empirical study of Nordic firms

Firms’ financing, boards of directors’ characteristics, investments, and firm-performance (financial and non-financial) occupy a pivotal place in corporate finance and corporate governance literature. The current study explores if causalities between the abovementioned four distinct albeit inter-related phenomena follow any pattern. The data comprising of 1240 firm-years belonging to Finland, Norway, Sweden, and Denmark for the period of 2003 to 2018 have been analyzed by applying multivariate linear regression and principal component analysis. The findings show that the impact of boards of directors’ characteristics is stronger on capital structure, however, weaker on investments and financial performance. The major contribution of the article is creating a set orderly and sequential causalities between financing, boards of directors’ characteristics, investments, and firm-performance.

Remaining Useful Life Prediction with Similarity Fusion of Multi-Parameter and Multi-Sample Based on the Vibration Signals of Diesel Generator Gearbox

The prediction of electrical machines’ Remaining Useful Life (RUL) can facilitate making electrical machine maintenance policies, which is important for improving their security and extending their life span. This paper proposes an RUL prediction model with similarity fusion of multi-parameter and multi-sample. Firstly, based on the time domain and frequency domain extraction of vibration signals, the performance damage indicator system of a gearbox is established to select the optimal damage indicators for RUL prediction. Low-pass filtering based on approximate entropy variance (Aev) is introduced in this process because of its stability. Secondly, this paper constructs Dynamic Time Warping Distance (DTWD) as a similarity measurement function, which belongs to the nonlinear dynamic programming algorithm. It performed better than the traditional Euclidean distance. Thirdly, based on DTWD, similarity fusion of multi-parameter and multi-sample methods is proposed here to achieve RUL prediction. Next, the performance evaluation indicator Q is adopted to evaluate the RUL prediction accuracy of different methods. Finally, the proposed method is verified by experiments, and the Multivariable Support Vector Machine (MSVM) and Principal Component Analysis (PCA) are introduced for comparative studies. The results show that the Mean Absolute Percentage Error (MAPE) of the similarity fusion of multi-parameter and multi-sample methods proposed here is below 14%, which is lower than MSVM’s and PCA’s. Additionally, the RUL prediction based on the DTWD function in multi-sample similarity fusion exhibits the best accuracy.