Divisive Hierarchical Clustering Research Articles

Trait‐based analyses reveal global patterns in diverse diets of albacore tuna (Thunnus alalunga)

AbstractSimplifying complex species interactions can facilitate tracking and predicting functional responses to ecological stressors. This is important for highly migratory pelagic predators, exploiting diverse prey fields as they respond to dynamic environments. We reconstructed the historical resource use of albacore tuna (Thunnus alalunga) globally from the 1900s to 2015 and confirmed highly biodiverse diets with 308 prey species, and an additional 238 taxa at coarser taxonomic resolution. We synthesised prey diversity into seven functional trait guilds using hierarchical divisive clustering algorithms as a function of four traits that describe habitat use and influence predator–prey encounter rates – prey habitat association vertically in the water column, horizontally along the coastal to pelagic gradient, seasonal and diel vertical migratory behaviour. We explored variability in historical composition of albacore diets across geographies based on species identity, individual trait information and functional trait guilds using a multi‐matrix modelling framework. Taxonomic information remains important for trophic ecology, however, species‐based diet composition in albacore tuna was highly variable across geographies and years sampled, making interpretation of these patterns difficult. By simplifying species identity into habitat‐based traits, we highlight changes in prey resources consumed, such as the historical importance of near‐surface epipelagic prey resources from coastal to oceanic habitats, and seasonally migrating continental shelf prey, with less frequent pulses of deeper water and demersal taxa. Trait information and trait guilds serve as useful classification frameworks for identifying functionally redundant food web linkages across biodiverse prey, and will prove useful in tracking predators' foraging responses to changing resource availability.

Towards a personalized health care using a divisive hierarchical clustering approach for comorbidity and the prediction of conditioned group risks

The objective was to assess risk of hospitalization and mortality of comorbidities using divisive hierarchical risk clustering to advice clinical interventions. Subjects and Methods: Data from the EHR of a general population, 3799885 adults, followed by 5 years. Model were performed using Spark and Scikit-learn and accuracy for the models was analyzed. Results: The number of models generated depends in part on the number of chronic diseases included (ex testing a sample of six diseases, a total number of 397 models for all-cause mortality and 431 models for hospitalization). The estimated models offered an ordered selection for the relevant clinical variables and their estimated risk as a group and for the individual patient in the group. Accuracy was assessed according to age, sex and the cardinality of the comorbid groups. A mobile version and dashboard were developed. Conclusion: The software developed stratified hospital admission and mortality risk in clusters of chronic diseases, and for a given patient, it could advise intensifying treatment or reallocating the patient risk.

Crop Information Retrieval Framework Based on LDW-Ontology and SNM-BERT Techniques

Currently, on the Internet, the information about agriculture is augmenting extremely; thus, searching for precise, relevant data of various details is highly complicated. To deal with particular difficulties like lower relevancy rate, false detection of retrieval resources, poor similarity rate, unstructured data format, multivariate data, irrelevant spelling, and higher computation time, an intelligent Information Retrieval (IR) system is required. An IR Framework centered on Levenshtein Distance Weight-centric Ontology (LDW-Ontology) and Sutskever Nesterov Momentum-centred Bidirectional Encoder Representation from Transformer (SNM-BERT) methodologies is presented here to overcome the complications as mentioned earlier. Firstly, the data is pre-processed, transmuting the unstructured data into a structured format, thus mitigating the error probabilities. Then, the LDW-Crop Ontology construction is done regarding the structured data. In the methodology presented, significance, frequency,and the suggestion of word in mind are considered to build Crop ontology. In the MongoDB database, the data being constructed are amassed. Then, by utilizing SNM-BERT, the data is trained for IR regarding clustered input produced by Inter Quartile Pruning Range-centred Hierarchical Divisive Clustering (IQPR-HDC) model. The LDW is computed for the provided user query; subsequently, the similarity evaluation outcomes are obtained from the database. The experiential evaluation displays that when analogized with the prevailing methodologies, a better accuracy of 94 % for simple queries and 92% for complex queries is achieved. Along with retrieval rate with lower computation time is achieved by the proposed methodology.

Explainable semi-supervised AI for green performance evaluation of airport buildings

Despite the global sustainability trend, airport buildings have received limited attention regarding their environmental impact within the built environment and aviation sectors. With less than 1% of green building certifications worldwide, there is a lack of essential green performance assessment tools, including frameworks, primary datasets, and models that require less human supervision. Previous research has addressed these issues by proposing frameworks, primary datasets, and a supervised method that leverages Classification and Regression Trees. However, these methods still rely on human effort to fix credit scores and assign category weights, necessitating significant supervision from human experts. This study presents a green performance evaluation model for airport buildings using an explainable semi-supervised AI approach to reduce human dependency during the inference of green scores. Principal Component Analysis and Hierarchical Clustering algorithms (agglomerative and divisive) are employed to construct the models. The dataset's dimensionality is reduced based on metrics such as scatter coefficient, Psi Index, variation, and permutation. The cluster numbers are determined using elbow point and silhouette coefficient values, resulting in cluster initializations between 6 and 12. Performance measures, including the rand index, mean absolute error, precision, recall, F1-measure, and accuracy, are evaluated for the models. The AHC and DHC models exhibit a maximum accuracy of 74%, albeit with different cluster numbers. Consequently, models are developed with suitable cluster numbers to derive green rules with improved predictive accuracy. Each agglomerative and divisive hierarchical clustering model yields 12 practical green rules that can assist airport operators and facility managers in enhancing the green performance of their airport buildings. These explainable green rules serve as guidelines to improve airport facilities. Additionally, a web-based green rating tool is developed to demonstrate the proof-of-concept, utilizing the green rules extracted from the agglomerative model.

Fingerprinting Portland cements by means of 87Sr/86Sr and 143Nd/144Nd isotope ratios and geochemical profiles

This study uses conventional 87Sr/86Sr and 143Nd/144Nd isotope and interelement ratios of Ca, Sr, K, Mn, Mg and Ti as fingerprints for provenancing ordinary Portland cements (OPC). Herein, the first database of Sr and Nd isotope ratios investigated in OPCs, stemming from 29 cement plants located worldwide, was created. The results show that the Sr isotope ratios of OPCs are higher than those of seawater from the observed geological period. The spread of 143Nd/144Nd in OPCs is not as large as the spread for 87Sr/86Sr isotope ratios. However, the combination of both Sr and Nd isotope ratios provides the potential for distinguishing between cements of different production sites. Most of the OPCs investigated have measurable differences in their 87Sr/86Sr and 143Nd/144Nd isotope ratios, which can be employed as a valuable analytical fingerprinting tool. In the case of equivocal results, divisive hierarchical clustering was employed to help overcome this issue. The construction of geochemical profiles allowed the computing of suitably defined distances between cements and clustering them according to their chemical similarity. By applying this methodology, successful fingerprinting was achieved in 27 out of the 29 ordinary Portland cements that were analysed.

An Automatic Hierarchical Clustering Method for the LiDAR Point Cloud Segmentation of Buildings via Shape Classification and Outliers Reassignment

Shape segmentation in urban environments forms the foundation for tasks such as classification and reconstruction. Most artificial buildings with complex structures are composed of multiple simple geometric primitives. Based on this assumption, this paper proposes a divisive hierarchical clustering algorithm that uses shape classification and outliers reassignment to segment LiDAR point clouds in order to effectively identify the various shapes of structures that make up buildings. The proposed method adopts a coarse-to-fine strategy. Firstly, based on the geometric properties of different primitives in a Gaussian sphere space, coarse extraction is performed using Gaussian mapping and the DBSCAN algorithm to identify the primary structure of various shapes. Then, the error functions are constructed after parameterizing the recognized shapes. Finally, a minimum energy loss function is built by combining the error functions and binary integer programming (BIP) to redistribute the outlier points. Thereby, the accurate extraction of geometric primitives is achieved. Experimental evaluations on real point cloud datasets show that the indicators of precision, accuracy, and F1 score of our method are 0.98, 0.95, and 0.96 (point assignment) and 0.97, 0.95, and 0.95 (shape recognition), respectively. Compared with other state-of-the-art methods, the proposed method can efficiently segment planar and non-planar structures with higher quality from building point clouds.

Towards Creativity and Innovation in Universities: Study on Central and Eastern Europe

AbstractUniversities have a pivotal role to play in the challenging restructuring process that economies are undergoing due to higher education’s critical role in the information age. Universities can contribute to this process of change by creating new knowledge as well as by shaping new skills, which are necessary for exploiting modern technologies and understanding how technology interacts with society. The aim of this paper is to compare universities from Central and Eastern Europe (CEE) in terms of creativity related to research and educational activity. The knowledge triangle concept is employed to compare involvement of universities in education, research, and innovation. The divisive hierarchical clustering was used to group countries by knowledge triangle performance. Five clusters have been identified regarding research-education-innovation interactions. The study contributes by shedding light on education as the missing link between research and innovation in the CEE countries and extends knowledge about heterogeneity of this region in terms of creativity. It also contributes to the discourse how innovation policies can support universities’ structural transformations towards integration of their education, research, and innovation activities that will enable to confront an array of societal challenges.

Genome sequencing reveals molecular subgroups in oral epithelial dysplasia.

This study aimed to analyze the molecular characteristics of oral epithelial dysplasia (OED), highlighting the pathways and variants of genes that are frequently mutated in oral squamous cell carcinoma (OSCC) and other cancers. Ten archival OED cases were retrieved for retrospective clinicopathological analysis and exome sequencing. Comparative genomic analysis was performed between high-grade dysplasia (HGD) and low-grade dysplasia (LGD), focusing on 57 well-known cancer genes, of which 10 were previously described as the most mutated in OSCC. HGD cases had significantly more variants; however, a similar mutational landscape to OSCC was observed in both groups. CASP8+FAT1/HRAS, TP53, and miscellaneous molecular signatures were also present. FAT1 is the gene that is most affected by pathogenic variants. Hierarchical divisive clustering showed division between the two groups: "HGD-like cluster" with 4HGD and 2LGD and "LGD-like cluster" with 4 LGD. MLL4 pathogenic variants were exclusively in the "LGD-like cluster". TP53 was affected in one case of HGD; however, its pathway was usually altered. We describe new insights into the genetic basis of epithelial malignant transformation by genomic analysis, highlighting those associated with FAT1 and TP53. Some LGDs presented a similar mutational landscape to HGD after cluster analysis. Perhaps molecular alterations have not yet been reflected in histomorphology. The relative risk of malignant transformation in this molecular subgroup should be addressed in future studies.

An unsupervised machine learning based ground motion selection method for computationally efficient estimation of seismic fragility

AbstractIn the context of performance‐based earthquake engineering (PBEE), response‐history analysis is currently considered an analytical tool for developing fragility curves. Typically, this involves subjecting a structural system to a large number of ground motion records (GMRs) representing seismic hazards at a site of interest and may be a time‐consuming task. To address this computational challenge, this study proposes a method for selecting a representative subset of GMRs that enables the reproduction of the fragility curve of the general GMR set. In this method, dimension reduction techniques are used to preferentially extract the principal features of earthquake intensity measures, which are applied to construct the feature space. Then, the divisive hierarchical clustering technique is applied to the feature space to obtain a subset of GMRs from the general set until the fragility curve converges. The performance of the proposed method is successfully demonstrated through various numerical examples that include a wide class of single‐degree‐of‐freedom systems and two steel‐frame buildings. The results confirm that the seismic hazard at a given site represented by a general GMR set can be covered in structural fragility estimation using a representative subset of GMRs selected based on the proposed method. The proposed method could contribute to significantly reducing the computational costs for structural fragility estimation without compromising the accuracy.

Achieving Agility in Projects Through Hierarchical Divisive Clustering Algorithm

Achieving Agility in Projects Through Hierarchical Divisive Clustering Algorithm

Agglomerative and divisive hierarchical Bayesian clustering

Cluster analysis methods are designed to discover groups of subjects or objects in datasets by uncovering latent patterns in data. Two model-based Bayesian hierarchical clustering algorithms are presented—divisive and agglomerative—that return nested clustering configurations and provide guidance on the plausible number of clusters in a principled way. These algorithms outperform many existing clustering methods on benchmark data. The methods are applied to identify subpopulations among Parkinson's disease subjects using only baseline data, and differing patterns of progression in the few years following diagnosis are demonstrated in the identified clusters.

Network-Informed Constrained Divisive Pooled Testing Assignments.

Frequent universal testing in a finite population is an effective approach to preventing large infectious disease outbreaks. Yet when the target group has many constituents, this strategy can be cost prohibitive. One approach to alleviate the resource burden is to group multiple individual tests into one unit in order to determine if further tests at the individual level are necessary. This approach, referred to as a group testing or pooled testing, has received much attention in finding the minimum cost pooling strategy. Existing approaches, however, assume either independence or very simple dependence structures between individuals. This assumption ignores the fact that in the context of infectious diseases there is an underlying transmission network that connects individuals. We develop a constrained divisive hierarchical clustering algorithm that assigns individuals to pools based on the contact patterns between individuals. In a simulation study based on real networks, we show the benefits of using our proposed approach compared to random assignments even when the network is imperfectly measured and there is a high degree of missingness in the data.

Uncovering material deformations via machine learning combined with four-dimensional scanning transmission electron microscopy

Understanding lattice deformations is crucial in determining the properties of nanomaterials, which can become more prominent in future applications ranging from energy harvesting to electronic devices. However, it remains challenging to reveal unexpected deformations that crucially affect material properties across a large sample area. Here, we demonstrate a rapid and semi-automated unsupervised machine learning approach to uncover lattice deformations in materials. Our method utilizes divisive hierarchical clustering to automatically unveil multi-scale deformations in the entire sample flake from the diffraction data using four-dimensional scanning transmission electron microscopy (4D-STEM). Our approach overcomes the current barriers of large 4D data analysis without a priori knowledge of the sample. Using this purely data-driven analysis, we have uncovered different types of material deformations, such as strain, lattice distortion, bending contour, etc., which can significantly impact the band structure and subsequent performance of nanomaterials-based devices. We envision that this data-driven procedure will provide insight into materials’ intrinsic structures and accelerate the discovery of materials.

A divisive hierarchical clustering methodology for enhancing the ensemble prediction power in large scale population studies: the ATHLOS project.

The online version contains supplementary material available at 10.1007/s13755-022-00171-1.

A Divisive Hierarchical Clustering Approach to Hyperspectral Band Selection

Band selection, considered as an effective dimensionality reduction technique for hyperspectral imagery (HSI), has become a hot topic for decades. Although various clustering-based methods have been applied to band selection, only a few studies explored the hierarchical structure among different spectral bands. And with regard to conventional hierarchical clustering, implemented in an agglomerative manner, both efficiency and accuracy of band selection still remain to rise. Moreover, the noise sensitivity is a defect inherent in the procedure of clustering. To address these issues, we propose a divisive hierarchical clustering approach (DHCA) to hyperspectral band selection. Inspired by divisive analysis, DHCA is designed to obtain any number of band subsets, which captures the intrinsic hierarchy of hyperspectral bands simultaneously. By introducing the local density into average dissimilarity, it can suppress the outliers clustering separately. Also, given the order of the spectrum, channel interval makes the similarity more rational among bands. Finally, we select a representative band in each cluster from the information viewpoint to ensure the band subset with a high quality. Extensive experiments on three real public HSI datasets fully validate the superiority of the proposed method against state-of-the-art competitors.

Cognitive Complexity Beyond Generalization: A Subjective Rating for the Human Comprehension

The cognitive complexity of a software determines the comprehension effort of a particular individual faces when designing, developing and maintaining a software. The comprehension level tends to be varied with each human resulting the cognitive complexity a subjective measurement. Expressing the cognitive complexity as a form of metric quantifies the comprehensibility as a generic value, which does not imply the subjectivity of human factor. This study elaborates the significance of expressing the cognitive complexity as a form of a subjective rating. The cognitive complexity rating has been pioneered with respect to the human and programming dependent factors related to human cognition. The Divisive hierarchical clustering algorithm has been used to train and predict the cognition rating per user. It has been clearly elaborated the subjectivity of the cognitive ratings over the quantitative and static complexity values of current cognitive and software complexity metrics. Thereby, the concept of cognition rates has been proposed as a preliminary step of determining and expressing the cognitive complexity.

Adaptive Resonance Theory-Based Topological Clustering With a Divisive Hierarchical Structure Capable of Continual Learning

Adaptive Resonance Theory (ART) is considered as an effective approach for realizing continual learning thanks to its ability to handle the plasticity-stability dilemma. In general, however, the clustering performance of ART-based algorithms strongly depends on the specification of a similarity threshold, i.e., a vigilance parameter, which is data-dependent and specified by hand. This paper proposes an ART-based topological clustering algorithm with a mechanism that automatically estimates a similarity threshold from the distribution of data points. In addition, for improving information extraction performance, a divisive hierarchical clustering algorithm capable of continual learning is proposed by introducing a hierarchical structure to the proposed algorithm. Experimental results demonstrate that the proposed algorithm has high clustering performance comparable with recently-proposed state-of-the-art hierarchical clustering algorithms.

Frailty phenotypes in patients with heart failure in the early post-discharge period: insights from the iCOR randomised controlled trial and a machine learning-based clustering analysis

Abstract Background/Introduction Several variables such as clinical, socioeconomic, functional or cognitive, among others can have an impact on the prognosis of heart failure (HF) patients despite the optimisation of follow-up strategies (e.g. telemedicine [TM] solutions). The clustering of HF patients may to identify different patient frailty phenotypes. Purpose The aim of this study was to perform a machine learning-based clustering analysis to identify different patient frailty phenotypes in a cohort of HF patients recruited in a randomized clinical trial (The Insuficiència Cardíaca Optimitzaciό Remota [iCOR] study). Methods We performed the clustering analysis on the basis of 8 frailty-related dimensions. To define the number of clusters, dissimilarity matrix was calculated with Gower's distance. Then, hierarchical divisive clustering was performed. Using then Elbow and Silhouette to analyse how the within sum of squares changes for the different number of clusters, the final number of clusters were chosen. The incidence proportion of the each of the study endpoints (non-fatal HF events as primary endpoint and all-cause hospitalization, all-cause death and the composite endpoint combining of all-cause death or non-fatal HF events as secondary endpoints) was calculated for cluster. Results 5 different frailty phenotypes were identified. Cluster 1 (29 patients, 16%) comprised patients with the best reported self-perceived health status (QoL), fair emotional-affective status, but low levels of self-care. Cluster 2 (41 patients, 23%) included the youngest patients with the highest level of education and a better level of cognition. Cluster 3 (68, 38%) encompassed the patients who had the best level of self-care behaviour (18.9±9.8), greater physical and instrumental functioning for activities of daily living (ADL) and a lower rate of comorbidities. Patients in the Cluster 4 (30 patients, 17%) tended to be elderly females with poor health-related QoL, and a higher level of functional dependence. Finally, Cluster 5 was the smallest group (10 patients, 6%), encompassing the oldest patients with low level of education, a worse affective-emotional state, a significant cognitive decline and a higher proportion of comorbidities compared to the other clusters. Cluster 4 had the highest incidence rate of the primary endpoint (57 per 100 patient-years at risk, 95% CI [37.4–74–5]) and a higher incidence of all-cause hospitalization and of the combined variable of all-cause of death or non-fatal HF events. Conclusion(s) Using the cluster analysis, we were able to stratify HF patients according to the stage of their impairment and vulnerability in each of the different frailty domains. This will allow clinicians to incorporate holistic multi-domain assessments in HF programmes to identify patients' needs and provide each patient with personalised and structured follow-up programme according to patient's needs (personalised and precision medicine). Funding Acknowledgement Type of funding sources: None. Radar chart to compare frailty clustersClinical endpoints according clusters

Clusters Across Multiple Domains of Health-Related Quality of Life Reveal Complex Patient Outcomes After Subarachnoid Hemorrhage.

Patients with subarachnoid hemorrhage were prospectively identified. Clinical data were prospectively recorded. Health-related quality of life was prospectively assessed at 3-month follow-up using the National Institutes of Health Patient Reported Outcomes Measurement Information System and neuro-quality of life in the domains of mobility, cognitive function, satisfaction with social roles, and depression. We used k-means clustering to analyze prospectively recorded health-related quality of life data, identifying clusters of complex patient outcomes. Decision tree analysis identified index hospital stay factors predictive of a patient having a particular complex patient outcome at follow-up. Academic medical center. One hundred three survivors of subarachnoid hemorrhage. None. We analyzed 103 patients, of whom 75 (72.8%) were female, and mean age was 53.6 ± 13.4 years. There were three complex patient outcomes: health-related quality of life greater than 1 sd better than the U.S. mean across all domains (n = 23, 22.3%), health-related quality of life greater than 1 sd worse than U.S. mean across all domains (n = 26, 25.2%), and satisfaction with social roles greater than 0.5 sd worse than U.S. mean with cognitive function, depression, and mobility scores near the U.S. mean (n = 54, 52.4%). In decision tree analysis, hospital disposition and Hunt and Hess Grade were associated with complex patient outcome. Complex patient outcomes across multiple domains of health-related quality of life at follow-up after subarachnoid hemorrhage are distinct and may be predictable.

Determination of cultural areas based on medieval pottery using an original divisive hierarchical clustering method with geographical constraint (MapClust)

Spatial representation based on the distribution of artefact, especially pottery, is widely used in archaeology. The questions raised are often related for the construction of economic or cultural areas, based on fabrics, style or types of vessels. The middle Loire Valley with the study of an important corpus of medieval pottery constitutes an essential framework for understanding this mechanisms over long time. The corpus of data collected and studied for over twenty years includes archaeological sites with reliable chrono-stratigraphic sets whose pottery assemblages are sufficiently large to respond to the problem posed.The permanent increase of the corpus of data, especially since the last publication on this subject in 2013 for which the approach was purely empirical, requires the development of statistical tools adapted here to the spatial clustering analysis of the data. This article deals with a divisive hierarchical clustering method with geographical constraints: MapClust. It is based on spatial indicators called spatial patches. This statistical tool allows the archaeologist to visualize and analyse geographical patterns easily. The comparison of the present results with those of 2013 is a way to see the relevance of the method.