Ward's Algorithm Research Articles

Cluster analysis applied to obtaining reference models for building thermal performance studies

Cluster analysis has often been used to obtain reference models. However, using different methods in each analysis step may yield different results. Furthermore, the validation of the results is often neglected. This study aims to develop a cluster analysis method to obtain reference models for use in thermal performance studies of buildings from different clustering configurations. The procedure proposed is composed of six steps: (1) initial database composition, (2) data matrices formation, (3) cluster analysis, (4) determination of reference models, (5) methods validation and (6) results interpretation. Different data treatments, similarity measures and partitioning algorithms were combined in the cluster analysis to define the method with the best cluster formation. The weighting factor was the data treatment that most contributed to obtaining suitable clustering solutions. City-block, Euclidean Distance and Squared Euclidean Distance similarity measures resulted in suitable formations as well as Complete Linkage, Ward and K-means algorithms. Two clusters were obtained from the cluster methodology, for which two reference models were determined. Hypothesis tests showed that clusters differ for most performance indicators. It was concluded that the procedure proposed could identify the combination of methods that results in the best application of cluster analysis. The main contribution of this paper is to introduce a procedure for validating clusters. This procedure is an applicable technique to obtain reference models.

Clustering analysis for the evolutionary relationships of SARS-CoV-2 strains.

To explore the differences and relationships between the available SARS-CoV-2 strains and predict the potential evolutionary direction of these strains, we employ the hierarchical clustering analysis to investigate the evolutionary relationships between the SARS-CoV-2 strains utilizing the genomic sequences collected in China till January 7, 2023. We encode the sequences of the existing SARS-CoV-2 strains into numerical data through k-mer algorithm, then propose four methods to select the representative sample from each type of strains to comprise the dataset for clustering analysis. Three hierarchical clustering algorithms named Ward-Euclidean, Ward-Jaccard, and Average-Euclidean are introduced through combing the Euclidean and Jaccard distance with the Ward and Average linkage clustering algorithms embedded in the OriginPro software. Experimental results reveal that BF.28, BE.1.1.1, BA.5.3, and BA.5.6.4 strains exhibit distinct characteristics which are not observed in other types of SARS-CoV-2 strains, suggesting their being the majority potential sources which the future SARS-CoV-2 strains' evolution from. Moreover, BA.2.75, CH.1.1, BA.2, BA.5.1.3, BF.7, and B.1.1.214 strains demonstrate enhanced abilities in terms of immune evasion, transmissibility, and pathogenicity. Hence, closely monitoring the evolutionary trends of these strains is crucial to mitigate their impact on public health and society as far as possible.

Combination of bulk RNA sequencing and scRNA sequencing uncover the molecular characteristics of MAPK signaling in kidney renal clear cell carcinoma.

The MAPK signaling pathway significantly impacts cancer progression and resistance; however, its functions remain incompletely assessed across various cancers, particularly in kidney renal clear cell carcinoma (KIRC). Therefore, there is an urgent need for comprehensive pan-cancer investigations of MAPK signaling, particularly within the context of KIRC. In this research, we obtained TCGA pan-cancer multi-omics data and conducted a comprehensive analysis of the genomic and transcriptomic characteristics of the MAPK signaling pathway. For in-depth investigation in KIRC, status of MAPK pathway was quantitatively estimated by ssGSEA and Ward algorithm was utilized for cluster analysis. Molecular characteristics and clinical prognoses of KIRC patients with distinct MAPK activities were comprehensively explored using a series of bioinformatics algorithms. Subsequently, a combination of LASSO and COX regression analyses were utilized sequentially to construct a MAPK-related signature to help identify the risk level of each sample. Patients in the C1 subtype exhibited relatively higher levels of MAPK signaling activity, which were associated with abundant immune cell infiltration and favorable clinical outcomes. Single-cell RNA sequencing (scRNA-seq) analysis of KIRC samples identified seven distinct cell types, and endothelial cells in tumor tissues had obviously higher MAPK scores than normal tissues. The immunohistochemistry results indicated the reduced expression levels of PAPSS1, MAP3K11, and SPRED1 in KIRC samples. In conclusion, our study represents the first integration of bulk RNA sequencing and single-cell RNA sequencing to elucidate the molecular characteristics of MAPK signaling in KIRC, providing a solid foundation for precision oncology.

An Examination of Aggregate Side-Effects Mapped to Therapies Experienced By Patients with Relapsed Large B-Cell Lymphomas - an Analysis from the Lymphoma Coalition's 2022 Global Patient Survey

Introduction The identification of side-effects associated with a particular therapy relies on univariable assessments of incidence across treatment arms. While this approach serves to set the groundwork for the establishment of causation, it is suboptimal when the incidence is small or when novel therapeutics with limited deployment are being examined. In this study, we used a multivariate approach to aggregate multiple symptoms in order to segment our sample into those who had differing degrees of side-effect incidence. We used the resultant strata to examine the use of various therapies. Methods Patients with relapsed large B-cell lymphomas who responded to the Lymphoma Coalition's 2022 Global Patient Survey (N=215) were asked if they had ever been treated with a particular treatment (N = 20) and what side effects (N=32) they have ever experienced. Unsupervised hierarchical clustering utilizing Ward's algorithm was used to aggregate the incidence of side-effects (N = 32). Resultant strata from the hierarchical clustering procedure were subjected to contingency analysis with the use of 20 treatments. Differences were assessed with odds ratio's, 95% confidence intervals and p-values from the likelihood chi-square statistic as appropriate. Results The total number of respondents meeting the criteria for inclusion in this study was 215. The median age of respondents was 58 and ranged from 23-85 years. Females comprised 55% of the study sample. Figure 1 Illustrates the two-way clustering heat map. A subjective decision was made to segment the study sample at the first branch yielding two clusters: 1 (N = 76) and 2 (N = 139). The rationale for selecting two clusters was based upon an obvious difference in side-effect incidence. The cumulative incidence of side effects in Cluster 1 = 1570 yielding an incidence rate per patient of 11.3 which is contrasted with Cluster 2 having a cumulative side-effect incidence of 272 yielding a rate per patient of 3.6. Table 1 illustrates the odds ratios and 95% confidence intervals derived from contingency analysis utilizing Cluster as the explanatory variable. Significantly increased odds of CAR-T (p<0.0001), steroids (0.006) and radiation therapy (p = 0.01) use were found in Cluster 1 relative to Cluster 2. We were unable to detect a significant change in the odds ratio for the other 17 treatments. Conclusions These results suggest that there is an association of increased side effects for those receiving CAR-T. Additionally, radiotherapy was also found to be marginally more used in those with more side-effects. The fact that steroid use was also increased in this cluster of patients makes sense since that is a routine prophylaxis strategy for patients experiencing substantial side effects. From the heat map, we can infer two things. First, there are groups of side effects that tend to cluster together which may reflect the treatment heterogeneity provided to this population. Second, it appears that about one-third of these patients experience very few side effects which supports the idea of patient heterogeneity. There are some limitations to this work which includes the use of patient reported outcomes and the fact that we were unable to explicitly map the treatment course of every individual patient. However, this study does support the idea that the diversity of the side-effect profile for CAR-T may extend well beyond the traditional phenomena of cytokine release syndrome, blood, and neurological effects. We suspect that as the use of CAR-T expands, additional side-effects will become apparent. Future studies should focus on identifying individuals who are likely to experience significant side effects and map treatment strategies to the appropriate patients.

Assessing the security status and future scenarios of the Mediterranean region through the water-energy-food nexus

Over the past decades, the Mediterranean region has faced significant challenges due to the impacts of climate change and ongoing conflicts. This study proposes an assessment of the region's security status and potential future scenarios through the lens of the water-energy-food nexus, utilising indicators that align with the Sustainable Development Goals (SDGs). These indicators include agricultural yields, value-added, and land variations, water and sanitation services, income and inequality, use of renewable energy, carbon footprints, and political stability. To evaluate the situation, this analysis applies Ward's hierarchical clustering algorithm to group countries based on these indicators, examining the average terms and the years 2006 and 2015 for comparative analysis. Additionally, an exponential smoothing algorithm forecasts future trends and generates clusters for the years 2030 and 2050. By computing an index of convergence for each cluster and indicator, this contribution identifies areas of particular interest from a security perspective. The findings of this analysis reveal a growing polarisation within the Mediterranean region, with European countries and Israel forming one distinct group, and African and Eastern countries (excluding Israel) forming another. Notably, recurring disparities exist in variables such as agricultural land, political stability, violence, income per capita, and agricultural value added. Conversely, certain variables, including the Gini coefficient, prevalence of overweight population, and access to drinking water services, show signs of convergence. These results shed light on potential areas of both conflict and cooperation in the Mediterranean region, highlighting the importance of addressing the challenges posed by climate change. By understanding the geopolitical dynamics and identifying key areas of concern, policymakers can develop informed strategies to promote stability and sustainable development in the region.

A cluster analysis on the potential of livestock farming: Türkiye in a global context.

This study aims to rank, and cluster countries in terms of some essential livestock indicators, and evaluate the subject with specific attention to the conditions existing in Türkiye. The study material covers 142 countries whose 2020 data were fully reported by the World Food and Agriculture Organization, the World Trade Organization, and the World Bank. It was determined that the most appropriate number of clusters would be 4 and significant differences were found between these clusters in terms of medium and high effect size for all indicators (P < 0.01). The cluster of 31 countries, according to the Ward algorithm, and the cluster of 37 countries according to the K-Means algorithm, both including Türkiye, were found to have higher averages than the other clusters in terms of related indicators (P < 0.01). The common elements in both clusters were determined: Türkiye, USA, Germany, Argentina, Brazil, Ethiopia, Morocco, France, India, the Netherlands, England, Spain, Italy, Canada, Kazakhstan, Kenya, Colombia, Pakistan, Peru, Poland, and Russia. Taking into account the geographical proximity and economic relations, Türkiye should examine the livestock models followed by these countries and evaluate the possible areas of cooperation that can be developed and competition that could be faced by these countries.

Time-resolved laser-induced fluorescence spectroscopy and chemometrics for fast identification of U(VI)-bearing minerals in a mining context

We evaluated the potential of time-resolved laser-induced fluorescence spectroscopy (TRLFS) combined with chemometric methods for fast identification of U(VI)-bearing minerals in a mining context. We analyzed a sample set which was representative of several environmental conditions. The set consisted of 80 uranium-bearing samples related to mining operations, including natural minerals, minerals with uranium sorbed on the surface, and synthetic phases prepared and characterized specifically for this study. The TRLF spectra were processed using the Ward algorithm and the K-nearest neighbors (KNN) method to reveal similarities between samples and to rapidly identify the uranium-bearing phase and the associated mineralogical family. The predictive models were validated on an independent dataset, and then applied to test samples mostly taken from U mill tailings. Identification results were found to be in accordance with the available characterization data from X-ray diffraction (XRD) and scanning electron microscopy-energy dispersive X-ray spectrometry (SEM-EDX). This work shows that TRLFS can be an effective decision-making tool for environmental investigations or geological prospection, considering the large diversity of uranium-bearing mineral phases and their low concentration in environmental samples.

Typology of rabbit rearing systems in tropical areas: A case of Benin (West Africa)

The breeding of rabbits in tropical areas is confronted with enormous difficulties related to feeding and health, which limit the expansion and viability of these farms. The objective of this study is to carry out a typology of rabbit farms in tropical conditions to characterize the structure and function of this rearing in order to better understand their production results. A sample of 600 rabbit farms located throughout Benin was selected. The typology was carried out using multiple correspondence analysis (MCA) followed by Hierarchical Cluster Analysis (HCA) according to the Ward's aggregation algorithm and Euclidean distance, which identified five typological groups. Group 1 covered 45.7% of the farms and included small-scale production (fewer than 20 does) by professional breeders using traditional parasite control methods. Group 2 covered 3.3% of the rearing and includes a larger number of semi-extensive farms using self-produced feed. Group 3 (14.7%) was characterized by farms with fewer than 20 does that were semi-extensive and used more phytotherapy. In Group 4 (9.7% of the farms), the extensive mode was the most used, with veterinary medicine being the most used. Group 5 concentrated 26.7% of the farms and was marked by semi-extensive breeding. These farms did not record any cases of parasitosis. The typology carried out allowed a better understanding of the operating modes of these farms, their problems and the main limiting factors.

Clinical phenotypes of older adults with non-valvular atrial fibrillation not treated with oral anticoagulants by hierarchical cluster analysis in the ANAFIE Registry.

Although anticoagulants are indicated for many elderly patients with non-valvular atrial fibrillation (NVAF), some patients do not receive anticoagulant therapy, whose characteristics and outcomes are diverse. In this sub-analysis of the All Nippon AF In the Elderly (ANAFIE) Registry, the phenotypes of patients who were not receiving anticoagulants at baseline were evaluated by cluster analysis using Ward's linkage hierarchical algorithm. Of 32,275 enrolled patients, 2445 (7.6%) were not receiving anticoagulants. Two clusters were identified: (1) elderly paroxysmal AF (PAF) patients with a high proportion of catheter ablation history (57%) and (2) very elderly patients with a high prevalence of previous major bleeding (43%). Respective mean ages were 80.9 and 84.2 years, mean CHA2DS2-VASc scores were 3.8 and 4.9, PAF prevalences were 100.0% and 31.4%, proportions of patients with catheter ablation history were 21.0% and 7.9%, and proportions of patients with a history of major bleeding were 4.0% and 10.8%. Annual incidence rates were 2.72% and 8.81% for all-cause death, 1.66% and 5.85% for major adverse cardiovascular or neurological events, 1.08% and 3.30% for stroke or systemic embolism, and 0.69% and 1.19% for major bleeding, respectively. In this cohort of elderly NVAF patients from the ANAFIE Registry who were not receiving anticoagulants, over half had PAF with a high proportion of catheter ablation history and a low incidence of adverse outcomes; for them, non-prescription of anticoagulants may be partially understandable, but they should be carefully monitored regarding AF burden or atrial cardiomyopathy and be adequately anticoagulated when adverse findings are detected. The remaining were very elderly patients with a high prevalence of previous major bleeding and a high incidence of adverse outcomes; for them, non-prescription of anticoagulants is inappropriate because of the high thromboembolic risk. Registration: http://www.umin.ac.jp/; Unique identifier: UMIN000024006.

Clinical phenotypes of older adults with non-valvular atrial fibrillation not treated with oral anticoagulants by hierarchical cluster analysis in the ANAFIE registry

Abstract Background and purpose The All Nippon AF In the Elderly (ANAFIE) registry evaluated the current status of anticoagulant therapy in older adult patients (aged ≥75 years) with non-valvular atrial fibrillation (NVAF) in Japan. Although older adult NVAF patients requires anticoagulation therapy, some of them do not receive it and the reasons would be diverse. Therefore, we aimed to identify the phenotypes of older adult NVAF patients not treated with oral anticoagulants using data from the ANAFIE registry. Methods In this sub-analysis of the ANAFIE registry, the phenotypes of patients who were not receiving anticoagulants at baseline were evaluated by cluster analysis using Ward's linkage hierarchical algorithm. We used 20 categorical variables and 6 continuous variables for the cluster analysis which were used as the risk factors in the main analysis of the ANAFIE registry [1]. Results Of 33,275 enrolled patients, 2445 (7.3%) were not receiving anticoagulants. Two clusters were identified: (1) elderly paroxysmal AF (PAF) patients with a low prevalence of comorbidities (58%) and (2) very elderly patients with a high previous major bleeding prevalence (42%). For each cluster, mean ages were 79.9 and 85.7 years, mean CHADS2 scores were 2.3 and 3.1, PAF prevalences were 83% and 54%, heart failure prevalences were 18% and 48%, the proportions of patients with a history of major bleeding were 0% and 25%, the proportions of patients with a history of catheter ablation were 21% and 8%, respectively. Annual incidence rates of each cluster were 2.62% and 9.08% for all-cause death, 1.72% and 5.83% for major adverse cardiovascular or neurological events, 1.22% and 3.14% for stroke or systemic embolism, and 0.53% and 1.43% for major bleeding. Conclusions In this cohort of elderly NVAF patients not receiving anticoagulants, more than half (∼60%) were PAF patients and had a low incidence of adverse outcomes. The remaining ∼40% were characterized as very elderly patients with a high prevalence of previous major bleeding and a high incidence of adverse outcomes. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): Daiichi Sankyo Co., Ltd.

Expeditious detection of Fusarium graminearum infection in rice by FTNIR using hierarchical cluster analysis

The lurking appositeness of Fourier transform near infrared spectroscopy (FTNIR) as a screening tool with ward's algorithm was implemented for rice samples spoiled by Fusarium graminearum followed by hierarchical cluster analysis (HCA) and histograms with storage days (0–35). The spectral inspection of F. graminearum spoiled rice was done at wave numbers ranging from 12,000 to 4000 cm−1 and the HCA was executed using vector normalization as a pre-processing method with a Euclidean distance of 0.017 with a wavenumber of 3595–12489 cm−1. The dissimilarities’ between rice spoiled by F. graminearum were computed using Pearson's correlation coefficients, which were further converted to D values. The major absorbance bands were detected at 7220.15, 6629.042, 5968.17, and 4981.861 cm−1. Further, it was observed that the absorbance units (AU) increased with storage days, the concentration of F. graminearum, and the moisture content (MC) of the rice grains. A total of 10 classes were constituted with a ‘D’ value of 36-153 for 0-35 storage days. The width of each class varied between 0.04 and 0.13, with an X mean D value of 0.37–1.31. The results further uncovered that the % accuracy of classification for 0-14 days was achieved at 100%, and 87.58% for 14–35 days up to 1 standard deviation (SD) respectively.

EICOSANOIDS AS HALLMARKS OF CHRONIC MYELOID LEUKEMIA PROGRESSION AND RESISTANCE TO TKI

Chronic myeloid leukemia (CML) is a Ph-positive myeloproliferative neoplasm giving rise to BCR-ABL1 oncogene. Usually, the disease is diagnosed in the chronic phase (CML-CP) with uncontrolled clonal expansion and high white blood cell count. Without therapy intervention, CML progress to advanced phases (accelerated and blastic phases; CML-AP). The Tyrosine Kinase Inhibitors (TKI) are the first-line therapeutic approach for CML. TKI targets the BCR-ABL tyrosine kinase activity and restrain the disease. The knowledge about the mechanisms involved in disease progression and resistance to TKI, beyond BCR-ABL1 , is limited. The relation among TKI treatment, disease progression and eicosanoids pattern in CML is unknown. Lipid mediators were quantified in healthy individuals (Controls; n = 9) as well as in CML patients in CP (n = 10), CML-AP (n = 13) and CML patients in molecular remission post-TKI therapy (RM; n = 6) using Liquid Chromatography/Mass Spectrometry (LC/MS). Mass spectral data were acquired with negative electrospray ionization. ProteinWizard was used to convert files to be used in bioinformatic subsequent analysis. Mummichog software was used for metabolic pathway enrichment analysis. Heatmaps were generated with the R package gplots and hierarchical clustering was performed with the R package amap with Spearman distance method and ward linkage algorithm. Volcano and bubble plots were obtained with the package ggplot2. We described the metabolomic profiles of CML in different phases and post-TKI therapy. The metabolic patterns were able to distinguish the CML patients in different phases and resistant to TKI. The metabolomic results suggested an association between reprogramming in lipid metabolism and CML progression and resistance to TKI therapy. CML patients showed to be a lower producer of some eicosanoids compared to controls. Interestingly, AP-CML produced less 12-HETE and EPA than CP-CML patients. CML patients in molecular remission produced more EPA in comparison with each other group. Taken together, the results established the eicosanoid profile of CML patients who progress to advanced phases and fail to respond to TKI.

New approach for hair keratin characterization: Use of the confocal Raman spectroscopy to assess the effect of thermal stress on human hair fibre.

ObjectiveThe objective of our research was to investigate the heat‐protecting effect of a product ex vivo and in vivo on human hair fibres.MethodsA preparatory study was carried out in order to determine an optimal threshold of thermal stress. For this, the structure of cross‐sections of the hair fibre was observed by optical microscopy. Then, Scanning Electron Microscopy (SEM) and Confocal Raman Spectroscopy (CRS) were applied to analyse ex vivo and in vivo morphological and molecular damage in hair structure after heat stress. Finally, in vivo tests were used to collect consumer perception.ResultsThe preparatory study enabled us to determine an optimal stress threshold of 10 heating cycle for SEM and 5 heating cycle for CRS. Based on spectral hierarchical classification using Ward's clustering algorithm, the ex vivo Raman results show that the spectral signature of the hair treated and heated is very close to the negative control. This shows that the product preserves the keratin structure after thermal stress. These results were also confirmed by an in vivo Raman analysis performed on hair samples from 5 donors. In concordance with Raman results, SEM shows that treated hair presents lesser “bubbles” and “crackling” on the hair surface. Finally, the in vivo studies proved that hair was more protected from heat.ConclusionThe authors concluded that the product shows protective properties with respect to morphological and molecular heat damage. We also demonstrate that the product promotes the α‐helix keratin conformation and preserves the S‐S disulfide bands.

Genetic diversity in reproductive traits of Braunvieh cattle determined with SNP markers.

Braunvieh is an important dual‐purpose breed in the Mexican tropics. The study of its genetic diversity is key to implementing genetic improvement programs. This study was conducted to determine genetic diversity of reproductive traits in a Mexican Braunvieh beef cattle population using single nucleotide polymorphisms in candidate genes. Information from 24 genes with 52 intra‐genic loci reported in literature to be associated with productive life, pregnancy rate and cow and heifer conception rate of 150 Braunvieh males and females was considered. Observed heterozygosity (Ho) revealed high genetic diversity for the studied traits, Ho = 0.42 ± 0.087, relative to that of other populations of the same breed. Cluster analyses were carried out using the Ward and K‐means algorithms. These analyses revealed high genetic diversity that was observed in the biplot of non‐metric multi‐dimensional scaling. It was found that clustering strategy allowed visualisation of distant groups by genotype but not by favourable alleles in all the loci. We found that the genes CSNK1E, DNAH11, DSC2, IBSP and OCLN affected most of the traits in our study and they were highly informative. Therefore, they represent a potential resource for selection and crossbreeding programs of the traits studied in Braunvieh. The analyses showed that the Mexican Braunvieh population has a high level of genetic diversity, arguably due to decades‐long adaptation to the Mexican tropics.

Understanding tourist behaviour towards destination selection based on social media information: an evaluation using unsupervised clustering algorithms

PurposeThe study aims to cluster the travellers based on their social media interactions as well as to find the different segments with similar and dissimilar categories according to traveller's choice. The study also aims to understand the behaviour of clusters of the travellers towards destination selection and accordingly make the tour packages in order to improve tourists' satisfaction and gain viable benefits.Design/methodology/approachAgglomerative hierarchical clustering with Ward's minimum variance linkage algorithm and model-based clustering with parameterized finite Gaussian mixture models has been implemented to achieve the respective goals. The dimension reduction (DR) technique was introduced for better visualizing clustering structure obtained from a finite mixture of Gaussian densities.FindingsA total of 980 travellers have been clustered into 8 different interest groups according to their tourism destinations selection across East Asia based on individual social media feedback. For selecting the optimal number of clusters as well as the behaviour of the interested travellers groups, both these proposed methods have shown remarkable similarities. DR technique ensures the reduction in dimensionality with seven directions, of which the first two directions explained 95% of total variability.Practical implicationsTourism organizations focus on marketing efforts to promote the most attractive benefits to the clusters of travellers. By segmenting travellers of East Asia into homogeneous groups, it is feasible to choose a similar area to test different marketing techniques. Finally, it can be identified to which segments, new respondents or potential clients belong; consequently, the tourism organizations can design the tour packages.Originality/valueThe study has uniqueness in two aspects. Firstly, the study empirically revealed tourists' experience and behavioural intention to select tourism destinations and secondly, it finds quantifiable insights into the tourism phenomenon in East Asia, which helps tourism organizations to understand the buying behaviours of tourists' segments. Finally, the application of clustering algorithms to achieve the purpose of this study and the findings are very new in the literature on tourism, to understand the tourist behaviour towards destination selection based on social media reviews.

Temporal aggregation of time series to identify typical hourly electricity system states: A systematic assessment of relevant cluster algorithms

Comprehensive numerical models are pivotal to analyze the decarbonization of electricity systems. However, increasing system complexity and limited computational resources impose restrictions to model-based analyses. One way to reduce computational burden is to use a minimum, yet representative, set of system states for model simulation. These states characterize fluctuating renewable generation and variable demand for electricity prevailing at a certain point in time. A review of possible time series aggregation techniques identifies cluster algorithms as most adequate, with k-Means and the Ward algorithm predominating. However, throughout the surveyed literature, the line of reasoning for the selection of these algorithms remains unclear. To support the electricity system modeling community in selecting an algorithm, this paper devises a systematic multi-stage evaluation approach to compare a large variety of cluster analysis configurations, differing in algorithm, cluster representation, and number of clusters. Results show that electricity demand and renewable energy generation time series can be compressed to below one percent while sustaining global characteristics of the original data. Two potent cluster configurations are identified, confirming k-Means and WARD as being prevalent. Beyond electricity market data, the methodology can be applied to various types of fundamental time-dependent input data.

Differentiation of oil samples by isomeric composition of paraffins using cluster analysis methods

A new method is proposed for controlling the sources of oil samples by chemical composition for their differentiation, identification and establishment of their origin (specific wells of the operating field) using the agglomerative hierarchical data clustering and applying the J. Ward algorithm. The initial information is the data on the composition and relative contents of n-paraffins С17Н36 ÷ С34Н70, isoprenoids of pristane С19Н40 and phytane С20Н42 in oil samples obtained by selected-ion monitoring gas chromatography/mass spectrometry. The solution to the problem was carried out in the absence of the information about the magnitude of the total uncertainty in the data, without using the standard samples for their assessment and without the preliminary assumption of the number of expected clusters. The proposed technique’s possibilities were illustrated by the results of 20 × 27 data matrix processing based on the relative paraffin content in 27 encrypted anonymous samples. As an outcome, three clusters were identified, each of which included objects sampled from the specific well. Differentiating the sources of the samples using the traditional methods by comparing the amounts of n-paraffins and isoprenoids proved to be impossible. Due to the inevitable errors in the analysis results and the possible changes in the composition of the recovered oil, the values of the relative contents of paraffins in the oil samples from all three wells overlap and the sources remain indistinguishable. The newly proposed method made it possible to solve the differentiation problem of oil samples in organic geochemistry and environment control which are complicated by the wide variability of the composition of oil samples, the lability of petroleum system under the influence of natural and technogenic factors and the absence of the estimate of the actual errors in the data.

Application of Clustering in the Dimensionality Reduction Algorithms for Separation of Financial Status of Commercial Banks in Ukraine

The issue of determining the financial condition of commercial banks and separating investment-attractive banks from problem banks on this ground is extremely important for developing countries. The aim of this study is to make sure on the example of Ukraine that commercial banks really form separate clusters, where more reliable, stable and efficient banks are well separable from less successful ones in this regard. The study used the t-SNE and UMAP dimensionality reduction algorithms, and the Ward's Agglomerative Hierarchical Clustering algorithm. The results of visual analysis of two-dimensional t-SNE projections show that banks of different degrees of risk are well separable and have their own specifics. Clustering in the UMAP algorithm allowed distinguishing clusters with banks of Class A,

Research on Asset Allocation Based on Hierarchical Clustering and Fuzzy Theory

After Markowitz's portfolio theory was put forward in the 1950s, the optimization method of venture capital portfolio had made continuous progress in three branches. Among them, double objective function became a mainstream. On the basis of predecessors, this paper proposes an asset-weight-optimization method based on hierarchical clustering and fuzzy theory. Firstly, this paper extracts the four characteristics of stocks, uses the hierarchical clustering method to aggregate the stocks with strong correlation to form a group, and then selects two of the most representative stocks from the group (i.e. clustering the data points with the best performance in each group) to form a new stock portfolio. Then, the fuzzy programming of the expected return is carried out for the new portfolio. And finally, the optimized asset allocation weight is obtained. In the experiment, we use the data from CSMAR database for empirical analysis. For the data preprocessing process of hierarchical clustering, we choose Ward algorithm. We conduct experiments with stock pools in China with data sample sizes of 20, 35 and 45 respectively. Firstly, we use hierarchical clustering data preprocessing, then optimize to obtain the weight with fuzzy theory, and finally compare the performance of relevant indicators with different sample sizes. Experiments show that hierarchical clustering can reduce the dimension of stock correlation matrix, which breaks through the limitation of the number of stocks for experiments, realizing "large-scale stock selection". The performance of programming which uses hierarchical clustering and fuzzy theory method is better than direct linear programming, in terms of simulating the state of people's investment and meet people's expectations at the same time. The innovation of the method and the improvement of the performance of relevant index show that the method proposed in this paper has a certain reference.

The application of Uniform Manifold Approximation and Projection (UMAP) for unconstrained ordination and classification of biological indicators in aquatic ecology

The analysis of community structure in studies of freshwater ecology often requires the application of dimensionality reduction to process multivariate data. A high number of dimensions (number of taxa/environmental parameters × number of samples), nonlinear relationships, outliers, and high variability usually hinder the visualization and interpretation of multivariate datasets. Here, we proposed a new statistical design using Uniform Manifold Approximation and Projection (UMAP), and community partitioning using Louvain algorithms, to ordinate and classify the structure of aquatic biota in two-dimensional space. We present this approach with a demonstration of five previously published datasets for diatoms, macrophytes, chironomids (larval and subfossil), and fish. Principal Component Analysis (PCA) and Ward's clustering were also used to assess the comparability of the UMAP approach compared to traditional approaches for ordination and classification. The ordination of sampling sites in 2-dimensional space showed a much denser, and easier to interpret, grouping using the UMAP approach in comparison to PCA. The classification of community structure using the Louvain algorithm in UMAP ordinal space showed a high classification strength for data with a high number of dimensions than the cluster patterns obtained with the use of a Ward's algorithm in PCA. Environmental gradients, presented via heat maps, were overlayed with the ordination patterns of aquatic communities, confirming that the ordinations obtained by UMAP were ecologically meaningful. This is the first study that has applied a UMAP approach with classification using Louvain algorithms on ecological datasets. We show that the performance of local and global structures, as well as the number of clusters determined by the algorithm, make this approach more powerful than traditional approaches.