Ward's Clustering Method Research Articles

System Analysis and Reliability Analysis for Environmental Test Chambers

The purpose of this paper is to present the results of the application of various models to estimate the reliability of environmental test chambers, especially, the methodology proposed by the International Electrotechnical Commission (IEC), using the Homogeneous Poisson process (HPP) and Non-Homogeneous Poisson process (NHPP) models, is adopted first, and then, a non-monotonic trend test and bathtub curve intensity function not covered by IEC are used, and supplementary analysis is used to characterize the resulting failure intensity. For the first time, the stochastic process model was applied to evaluate the reliability of 20 environmental test chambers. The results show that the IEC standards process is suitable for the reliability evaluation of a single chamber, and 16 chambers conform to the HPP model and 4 chambers conform to the NHPP model. However, there is the power-law model (PLP) rejection cases among the overall description of multiple chambers by the IEC model. The rejected cases were analyzed again by using a non-monotonic trend test and constructing a double Weibull process tub curve strength function, and the 3-stage time interval of the bathtub curve failure is obtained, which is in line with the actual operation data. The Ward clustering method is used for the mean time between failures of 20 chambers, resulting in 4 types of chamber groups with different reliability values (71,52,100,130 days), which is of great significance for studying the reliability of the environmental test chamber and carrying out customized maintenance.

The determinants of COVID-19 case fatality rate (CFR) in the Italian regions and provinces: An analysis of environmental, demographic, and healthcare factors

The Italian government has been one of the most responsive to COVID-2019 emergency, through the adoption of quick and increasingly stringent measures to contain the outbreak. Despite this, Italy has suffered a huge human and social cost, especially in Lombardy. The aim of this paper is dual: i) first, to investigate the reasons of the case fatality rate (CFR) differences across Italian 20 regions and 107 provinces, using a multivariate OLS regression approach; and ii) second, to build a “taxonomy” of provinces with similar mortality risk of COVID-19, by using the Ward's hierarchical agglomerative clustering method. I considered health system metrics, environmental pollution, climatic conditions, demographic variables, and three ad hoc indexes that represent the health system saturation. The results showed that overall health care efficiency, physician density, and average temperature helped to reduce the CFR. By the contrary, population aged 70 and above, car and firm density, air pollutants concentrations (NO2, O3, PM10, and PM2.5), relative average humidity, COVID-19 prevalence, and all three indexes of health system saturation were positively associated with the CFR. Population density, social vertical integration, and altitude were not statistically significant. In particular, the risk of dying increases with age, as 90 years old and above had a three-fold greater risk than the 80–to–89 years old and four-fold greater risk than 70–to–79 years old. Moreover, the cluster analysis showed that the highest mortality risk was concentrated in the north of the country, while the lowest risk was associated with southern provinces. Finally, since prevalence and health system saturation indexes played the most important role in explaining the CFR variability, a significant part of the latter may have been caused by the massive stress of the Italian health system.

K-Means, Ward and Probabilistic Distance-Based Clustering Methods with Contiguity Constraint

We analyze some possibilities of using contiguity (neighbourhood) matrix as a constraint in the clustering made by the k-means and Ward methods as well as by an approach based on distances and probabilistic assignments aimed at obtaining a solution of the multi-facility location problem (MFLP). That is, some special two-stage algorithms being the kinds of clustering with relational constraint are proposed. They optimize division of set of objects into clusters respecting the requirement that neighbours have to belong to the same cluster. In the case of the probabilistic d-clustering, relevant modification of its target function is suggested and studied. Versatile simulation study and empirical analysis verify the practical efficiency of these methods. The quality of clustering is assessed on the basis of indices of homogeneity, heterogeneity and correctness of clusters as well as the silhouette index. Using these tools and similarity indices (Rand, Peirce and Sokal and Sneath), it was shown that the probabilistic d-clustering can produce better results than Ward’s algorithm. In comparison with the k-means approach, the probabilistic d-clustering—although gives rather similar results—is more robust to creation of trivial (of which empty) clusters and produces less diversified (in replications, in terms of correctness) results than k-means approach, i.e. is more predictable from the point of view of the clustering quality.

Panel estimation for the impact factors on carbon dioxide emissions: A new regional classification perspective in China

Global warming is an urgent issue facing humanity. Excessive emissions of CO2 are the leading cause of global climate change. As economy grows, China’s CO2 emissions grows rapidly correspondingly. Since 2006, China has ranked first in terms of total CO2 emissions, results of different contributions of different provinces. In view of the shortcomings in the previous regional classification, this work adopted a new regional classification framework which combines factor analysis and Ward clustering method, and classified 30 provinces into four regions by using panel data from 2002 to 2016. Then, this work investigated the regional differences in terms of the effect of population size, GDP per capita, energy structure, energy intensity, urbanization level and industrialization level on CO2 emissions by STIRPAT model. The outcomes indicated that population size, GDP per capita and energy intensity were the top three factors affecting CO2 emissions for all regions. The impact of population size on carbon emissions is not only determined by the size of the population. Along with the process of urbanization, the role of urbanization on CO2 emissions gradually changes from positive to negative. The factor urbanization had positive impacts in region (Ⅲ and Ⅳ) but negative impacts in region (Ⅰ and Ⅱ). The greatest positive impacts of energy structure can be seen in region IV. Yet, statistically, energy structure shows no significant impact in region Ⅱ. As for industrialization level, it demonstrated negative effect in region Ⅱ but positive effect in the other three regions. From these results, implications for CO2 emission-reduction policies in each region were discussed.

Precipitation trends in the island of Ireland using a dense, homogenized, observational dataset

AbstractA dense monthly precipitation dataset of Ireland and Northern Ireland was homogenized with several modern homogenization methods. The efficiency of these homogenizations was tested by examining the similarity of homogenization results both in the real data homogenization and in the homogenization of a simulated dataset. The analysis of homogenization results shows that the real dataset is characterized by a large number of, but mostly small, non‐climatic biases, and a moderate reduction of such biases can be achieved with homogenization. Finally, a combination of the ACMANT and Climatol homogenization results was applied to improve the data accuracy before the trend calculations. These two methods were selected for their proven high accuracy, missing data tolerance and ability to complete time series via the infilling of missing values before the trend calculations. Metadata were used within the Climatol method. To facilitate this analysis the study area was split into smaller climatic regions by using the Ward clustering method. Five climatic zones consistent with the known spatial patterns of precipitation in Ireland were established. Linear regression fitting and the Mann‐Kendall test were applied. Low frequency fluctuations were also examined by applying a Gaussian filter. The results show that the precipitation amount generally increases in the study area, particularly in the northwestern region. The most significant increasing trends for the whole study period (1941–2010) are found for late winter and spring precipitation, as well as for the annual totals. In the period from the early 1970s the increase of precipitation is general in all seasons of the year except in winter, but the statistical significance of this increase is weak.

Assessment of Turkish consumer attitudes using an Animal Welfare Attitude Scale (AWAS)

The aim of this study was to examine Turkish consumer attitudes towards animal welfare in terms of cognitive, affective and behavioral dimensions, using a bespoke Animal Welfare Attitude Scale (AWAS). An overall consumer attitude was also determined. The Delphi technique was used to establish an item pool to develop a questionnaire for the construction of the AWAS. This questionnaire was later used for data collection. A total of 2295 consumers were surveyed in 14 cities, in the 7 regions of Turkey. Descriptive statistics, exploratory factor analysis (EFA), confirmatory factor analysis (CFA), reliability analysis, Ward's hierarchical clustering method and One-way ANOVA were used to validate the questionnaire, and to analyze data. Results of the EFA allowed for allocation of 42 items collected under 3 dimensions (cognitive, affective and behavioral), that explained 72% of the total variance of the model. This factor structure was subsequently confirmed by a CFA performed on a different sample of 425 consumers. The Cronbach's Alpha coefficient for AWAS was calculated at 0.829. These results confirmed that the developed AWAS had a valid and reliable scale. The questionnaire showed that consumers' attitudes towards animal welfare were more negative at the behavioral dimension, than either at the cognitive or affective dimensions. Consumers in Turkey were ultimately divided into three groups according to their overall attitudes towards animal welfare as impassive, moderate or sensitive. One-third of Turkish consumers placed in the sensitive group, thus emphasizing a potential niche for animal-friendly food marketing in Turkey.

Clustering approach to urban rainfall time series prediction with support vector regression model

ABSTRACT The aim of rainfall prediction in urban areas is to provide information about precipitation, which may cause flooding resulting from the insufficient capacity of sewage systems. The research combined cluster analysis and support vector regression (CSVR) to create a simpler, and as effective as hybrid models, method predicting daily rainfall in Wroclaw, Poland. Results demonstrate that the clustering approach while modelling is reasonable, improves the quality of prediction and minimises error values. The best SVR performance, RMSE = 2.492 mm and r2 = 0.830 in the testing subset, was obtained for the Ward clustering method. Nevertheless, comparisons with MLP prediction, combined with k-means clustering, proved to be slightly more accurate and led to the creation of a model with RMSE = 1.974 and r2 = 0.899. The presented approach might constitute an alternative method to be used for creating rainfall patterns in stormwater discharge or management systems leading to an increase in flash flood safety in an urban catchment.

VARIATION AND RELATIONSHIPS OF AERIAL YIELD, MORPHO-AGRONOMIC TRAITS AND ESSENTIAL OIL COMPOSITION IN DOMESTIC POPULATIONS OF Ziziphora tenuior IN IRAN

The present study was carried out to evaluate aerial yield, agronomic traits, essential oil production, and their relationships in 21 populations of Ziziphora tenuior in Karaj, Iran over two years (2016–2017). The experimental layout was a randomized complete block design with three replications. Data were collected for GDD (Growth Degree Days), plant height, canopy area, stem number, flower number, aerial fresh yield, aerial dry matter (DM) yield, seed yield, essential oil percentage and oil production. The essential oil compositions were detected in accession 2929 (Tassoj) using GC/MS. Result of combined analysis of variance across two years showed significant differences between years, populations and population x year interaction for all traits (P < 0.01). In comparisons among populations, the population of Avaj and Ijroud with average values of 13.5 and 14.33 g/p had higher aerial dry yield and other agronomic traits followed by Meshkin2 and 3. For seed yield, Ormieh with average values of 3.11 g/p had higher production. For oil content, the populations of Tassoj, Salmas and Sharkord with average values of 1.25 to 1.56% had higher oil content and essential oil production. Result of correlation analysis showed that DM yield was positively correlated with GDD, plant height, canopy area, stem number, flower number. Oil content was negatively correlated with aerial DM yield. The results of PCA analysis showed that the first four components accounted for 50, 20, 12 and 10% of the total variation, respectively. Plant height, canopy area, stem number, flower number and aerial yield in the PCA1, essential oil content and oil production in the PCA2, seed yield in the PCA3 and GDD in PCA4 components were identified as important traits. Using Ward cluster method, the 21 populations were grouped into 3 clusters. In total, 23 compounds were identified in the essential oil from the aerial parts Z. tenuior. The major components in the essential oil were pulegone ranged (47.54 to 65.26%) followed by neo-iso-dihydro carveol, carvacrol, piperitenone, limonene, 1.8-cineole, p-mentha-3,8-dien, alpha-terpinene and terpinen-u-ol. It was concluded that there were wide range of variations for agronomic trait and essential oil in domestic germplasm of Z. tenuior to improved breeding new varieties in Iran.

What Ward's clustering method tells about the four largest emerging equity markets

Investing in stock market indices or ETFs could be more reasonable (and secure) for relatively new and incognisant investors who are personally unable to value of each stock–firm in a way. This paper attempts to group the well-known four largest emerging stock markets so-called BRIC or big-four economies namely Brazil, Russia, India and China based on return characteristics through Ward's hierarchical clustering method over the period of 2005 and 2015. Additionally, the first principle component (PCA) of the related indices is calculated and the abnormal variability is observed through control chart over time. Results indicate that Brazilian and Indian markets show more similarity over the studied period.

What Ward's clustering method tells about the four largest emerging equity markets

Investing in stock market indices or ETFs could be more reasonable (and secure) for relatively new and incognisant investors who are personally unable to value of each stock–firm in a way. This paper attempts to group the well-known four largest emerging stock markets so-called BRIC or big-four economies namely Brazil, Russia, India and China based on return characteristics through Ward's hierarchical clustering method over the period of 2005 and 2015. Additionally, the first principle component (PCA) of the related indices is calculated and the abnormal variability is observed through control chart over time. Results indicate that Brazilian and Indian markets show more similarity over the studied period.

Classification and ordination of plant communities in Abune Yosef mountain range, Ethiopia

Understanding how topography-soil-disturbance drives spatial distribution of vegetation is the interest of ecologists. This study was, conducted to investigate the topography-soil-disturbance and vegetation relationships in Abune Yosef mountain range, Ethiopia. A total of 85 nested sample plots measuring 400 m2, 25 m2 and 1 m2 were established for trees, shrubs and herbs respectively. Topographic, soil and disturbance variables were also assessed from each plot. Plant community classification was described and identified by agglomerative hierarchical clustering using Ward's minimum variance clustering methods. Shannon diversity index was employed to determine community diversity. After detecting the length of the first Detrended Correspondence Analysis (DCA) axis, Canonical Correspondence Analysis (CCA) forward and backward stepwise selection of environmental variables was performed based on their p-value by running permutation tests. The first axis explained 43.63% of the overall inertia and is correlated with Elevation, pH, slope aspect, total Nitrogen, soil organic Carbon & Clay. On the other hand, the second axis explained 32.06% of the total inertia and is correlated with bulk density, slope, logging, & available Phosphorus. The present study revealed that topographic variables have a profound influence on vegetation spatial distribution than soil and disturbance factors.

Cultural Differences and the Consumption Structure in the European Union Countries

The aim of the paper is to study the cultural differentiation in the European Union countries and compare it with the differentiation in their consumption spending. The question is whether similar countries in terms of culture have similar final consumption expenditures and consumption structures. Culture in this research is characterised by six Hofstede dimensions – power distance, individualism vs. collectivism, masculinity vs. femininity, uncertainty avoidance, long term orientation and indulgence vs. restraint. The consumption structure is characterised by share of durable goods, semi‑durable goods, non‑durable goods and services in households’ final consumption. In the analysis, the influence of culture on the share of non‑durable goods and services in final consumption expenditures is considered. Countries’ similarities in these two aspects are evaluated with the use of the cluster analysis approach – the k‑means algorithm and the Ward clustering method. The dependence between the structure of final consumption expenditures and culture is investigated using spatial autoregressive (SAR) and spatial error (SE) panel data models.

Modeling inactivation of Listeria monocytogenes, Shigella sonnei, Byssochlamys fulva and Saccharomycescerevisiae and ascorbic acid and β-carotene degradation kinetics in tangerine juice by pulsed-thermosonication

Inactivation of Listeria monocytogenes (serotype 4 ab), Shigella sonnei, Byssochlamys fulva and Saccharomyces cerevisiae and the degradation of β-carotene and ascorbic acid in tangerine juice subjected to pulsed-thermosonication were modeled using polynomial, log-logistic, Weibull, biphasic linear and modified Gompertz models. The models were assessed using the adjusted determination coefficient (adj-R2) and the root mean square error (RMSE). Then, Ward's clustering method was applied to classify the fitted models. The Weibull model showed correlation coefficients above 0.98. Moreover, a process performance criterion was developed to examine the simultaneous effect of inactivation of microorganisms and degradation of β-carotene and ascorbic acid. Results indicated the effect of temperature was higher than pulsed-ultrasound amplitude for L. monocytogenes, S. sonnei and S. cerevisiae, while the effect of these two parameters was similar for B. fulva. Regarding β-carotene and ascorbic acid, pulsed-sonication amplitude was the major parameter in decreasing theirs contents in juice. The treatment procedure in the optimal conditions had the highest effect on L. monocytogenes, while S. sonnei, S. cerevisiae and B. fulva were at similar levels. Therefore, the models obtained present a valuable tool to design and predict pasteurization processes of tangerine juice during pulsed-thermosonication.

National fiscal frameworks in the post-crisis European Union

Research background: The latest economic and financial crisis has seriously injured European Union Member States, affecting the condition of their public finances. In the face of the crisis, the EU made a special effort to increase the effectiveness of national fiscal frameworks, e.g. by improving the compliance with legislation. The post 2009 reforms were aimed at providing a solid economic foundation for the national fiscal frameworks, especially in the high-debt euro area countries.
 Purpose of the article: The goal of this research is twofold. Firstly, it aims to provide an outline of the national fiscal governance in the EU. Secondly, the paper analyzes the changes in the core measures of fiscal governance in the EU between the crisis period and the year 2016 (due to the latest available data) and investigates the similarities in the progress made by the 28 EU countries in restoring balance in public finance.
 Methods: To achieve the goal, the literature review and the analysis of core elements of national fiscal frameworks are provided. In the empirical section the grouping method for all 28 EU countries based on the Ward's agglomerative hierarchical clustering method is employed. The study uses data derived from the AMECO database (in the case of fiscal data) and the European Commission thematic data for quality indexes of particular elements of fiscal governance (numerical fiscal rules, medium-term budgetary frameworks and independent fiscal institutions).
 Findings & Value added: This paper contributes to the literature by, on the one hand, attempting to analyze changes in main fiscal governance measures and, on the other hand, by assessing their link with public finance through employment of the agglomerative clustering method. Based on the results, the conclusion about the importance of the improvement in fiscal frameworks is provided. The analysis shows that countries with better national fiscal framework achieved better results in public finances regardless the macroeconomic conditions.

Türkiye’de İllerin Gelişmişlik Düzeyi Dikkate Alınarak Sağlık Hizmetlerinin Kümeleme Analizi Aracılığıyla Değerlendirilmesi

Health is defined as a state of complete well-being and health services are such a kind of services which help to protect and sustain the health situation. There are several factors which affect health services and the level of services individuals utilized. One of the factors is that social and economic development level where individuals live. Turkey is one of the developing countries and despite this there are huge development differences among the cities. Differences could be seen in health services at the city level. Because of this reason, this study investigates the differences among the cities on the consideration of development and health service level. Ward clustering method is chosen for the analysis. The analysis consists of 81 cities and uses 31 health variables. Health variables are taken from “Health Statistics Annual” prepared by The Ministry of Health. Results reveal that there is a positive correlation between social-economic development and health service level for cities.

A Novel Classification and Identification Scheme of Emitter Signals Based on Ward's Clustering and Probabilistic Neural Networks with Correlation Analysis.

The rapid development of modern communication technology makes the identification of emitter signals more complicated. Based on Ward's clustering and probabilistic neural networks method with correlation analysis, an ensemble identification algorithm for mixed emitter signals is proposed in this paper. The algorithm mainly consists of two parts, one is the classification of signals and the other is the identification of signals. First, self-adaptive filtering and Fourier transform are used to obtain the frequency spectrum of the signals. Then, the Ward clustering method and some clustering validity indexes are used to determine the range of the optimal number of clusters. In order to narrow this scope and find the optimal number of classifications, a sufficient number of samples are selected in the vicinity of each class center to train probabilistic neural networks, which correspond to different number of classifications. Then, the classifier of the optimal probabilistic neural network is obtained by calculating the maximum value of classification validity index. Finally, the identification accuracy of the classifier is improved effectively by using the method of Bivariable correlation analysis. Simulation results also illustrate that the proposed algorithms can accurately identify the pulse emitter signals.

Morphological and molecular characterization of two distinct chilli cultivars from North Eastern India with special reference to pungency related genes

Abstract King chilli and Dalle Khursani are the two hottest chilli (Capsicum sp.) cultivars widely grown in NE region of India but a well-organized study pertaining to King chilli and Dalle Khursani is unavailable. We have characterized twenty-two accessions of King chilli and Dalle Khursani, collected from different states of NE region with respect to 27 morphological parameters and using 30 SSR markers. The allelic status of three key pungency related genes viz, putative aminotransferase (pAmt or Pun1), acyl-CoA synthetase (Acs1) and β-ketoacyl-ACP synthase (Kas1) was also determined. Maximum morphological variation was observed for fruit traits and traits like plant growth habit, plant height, etc., which can be used to distinguish between the two cultivars. SSR markers HpmsE062 and HpmsE063 were able to distinguish between the three different cultivars (including a nonpungent bell pepper), while HpmsE116 and HpmsE139 distinguish King chilli from Dalle Khursani. Ward clustering method grouped all the eleven Dalle Khursani accessions together with non-pungent C. annuum accession in one cluster. Sequencing across Acs1 revealed total of eight SNPs and three InDels. Seven SNPs including three within exonic regions were conserved between bell pepper, Dalle Khursani and the reference pungent C. annuum accession. A 2 bp InDel at position 50–51 bp and a 3 bp InDel at position 6313–6315 bp were found to be uniquely conserved haplotypes in both Dalle Khursani and King Chilli. A total of 16 SNPs were found within the sequenced region of Pun1 gene, of which eight SNPs lie within the coding region. Three novel exonic SNPs were found in Dalle Khursani. All eight SNPs found in the intronic region for Pun1 were conserved across Dalle Khursani and the two pungent reference sequences of C. annuum. Our SSR and SNP data suggests that Dalle Khursani is genetically closer to pungent C. annuum. This work provides insights into the genotype of two hottest chillies and paves way for their utilization in chilli improvement programme.

Functional approach in estimation of cultural ecosystem services of recreational areas

The article is devoted to the identification and analysis of cultural ecosystem services of recreational areas from the different forest plant functional groups in the suburbs of Kazan. The study explored two cultural ecosystem services supplied by forest plants by linking these services to different plant functional traits. Information on the functional traits of 76 plants occurring in the forest ecosystems of the investigated area was collected from reference books on the biological characteristics of plant species. Analysis of these species and traits with the Ward clustering method yielded four functional groups with different potentials for delivering ecosystem services. The results show that the contribution of species diversity to services can be characterized through the functional traits of plants. This proves that there is a stable relationship between biodiversity and the quality and quantity of ecosystem services. The proposed method can be extended to other types of services (regulating and supporting). The analysis can be used in the socio-economic assessment of natural ecosystems for recreation and other uses.

Using footprints to identify and sex giant pandas

Data on numbers and distribution of free-ranging giant panda are essential to the formulation of effective conservation strategies. There is still no ideal method to identify individuals and sex this species. The traditional bite-size method using bamboo fragments in their feces lacks accuracy. The modern DNA-based estimation is expensive and demands fresh samples. The lack of identifiable individual features on panda pelage and no apparent sexual dimorphism impede reliable estimation from camera trap images. Here, we propose an innovative and non-invasive technique to identify and sex this species using a footprint identification technique (FIT). It is based on a pairwise comparison of trails (unbroken series of footprints) using discriminant analysis, with a Ward's clustering method. We collected footprints from 30 captive animals to train our algorithm and used another 11 animals for model validation. The accuracy for individual identification was >90% for individuals with more than six footprints and 89% with fewer footprints per trail. The accuracy for sex discrimination was about 84% using a single footprint and 91% using trails. This cost-effective method provides a promising future for monitoring wild panda populations and understanding their dynamics and especially useful for monitoring reintroduced animals after the detachment of GPS collars. The data collection protocol is straightforward and accessible to citizen scientists and conservation professionals alike.

Quantifying the interactions of land management practices and agricultural productivity using a soil quality index

AbstractIt is desirable to develop an objective Soil Quality Index (SQI) to guide sustainable agronomic intensification, thereby promoting socio‐economic well‐being. This study pioneers the use of Ward's cluster and principal component regression methods to evaluate soil homogeneity and construct a SQI (expressed as %). Field data were acquired from five different sites within Ohio, USA, that were under no‐till (NT), conventional till (CT) management and natural vegetation (NV) land use. Soil pH, carbon/nitrogen (C/N) ratio, nitrate and soil organic carbon (SOC) concentrations were identified as primary drivers of soil quality. Based on Ward's cluster method, the soil properties of croplands were not significantly different from those under NV land use. However, SQI ranked surface soils under CT management as higher in quality than NV and NT managed soils, respectively. The coefficient of determination (R2) between SQI and corn (Zea mays L.) and soya bean [Glycine max (L.) Merr.] yields was 0.7 and 0.9, respectively, implying this SQI effectively relates soil properties, a function of anthropogenic land management practices, with crop yields. In future, time series analyses will be used to assess SQI versus crop yield dynamics, with key socio‐economic and climate variables.