Final Clustering Research Articles

Unified representation aggregated from graph and global features for multi-source data clustering

Multi-source data has received increasing attention due to its excellent performance in clustering tasks. However, existing multi-source data clustering methods utilize shallow graph learning methods to model similarity graphs of multi-source data, and ignore the importance of weak connections between samples within the same cluster. Thus, these connections fail to be explored in the global similarity graph of multi-source data. In this paper, we proposed a unified representation aggregated from graph and global features (UGGF) method for multi-source data clustering. Specifically, it utilizes the cross-source graph diffusion process to preserve invariant connections in multi-source similarity graphs and preserve weak connections between samples through their connection relationships between different sources. Furthermore, inspired by self-attention, the encoder of transformers is used to learn multi-source global feature representation. Then, global features and similarity graphs are integrated to comprehensively explore the multi-source data, aiming to obtain a unified representation for the final clustering task. The proposed UGGF method is validated on four benchmark datasets. Extensive experimental results demonstrate the superiority of the proposed method compared with other state-of-the-art multi-source data clustering methods.

A Novel and Effective Method to Directly Solve Spectral Clustering.

Spectral clustering has been attracting increasing attention due to its well-defined framework and excellent performance. However, most traditional spectral clustering methods consist of two separate steps: 1) Solving a relaxed optimization problem to learn the continuous clustering labels, and 2) Rounding the continuous clustering labels into discrete ones. The clustering results of the relax-and-discretize strategy inevitably result in information loss and unsatisfactory clustering performance. Moreover, the similarity matrix constructed from original data may not be optimal for clustering since data usually have noise and redundancy. To address these problems, we propose a novel and effective algorithm to directly optimize the original spectral clustering model, called Direct Spectral Clustering (DSC). We theoretically prove that the original spectral clustering model can be solved by simultaneously learning a weighted discrete indicator matrix and a structured similarity matrix whose connected components are equal to the number of clusters. Both of them can be used to directly obtain the final clustering results without any post-processing. Further, an effective iterative optimization algorithm is exploited to solve the proposed method. Extensive experiments performed on synthetic and real-world datasets demonstrate the superiority and effectiveness of the proposed method compared to the state-of-the-art algorithms.

Analyzing the Efficacy of K-Means Clustering and Logistic Regression For Diabetes Prediction

Diabetes causes a large number of deaths each year and a large number of people living with the disease do not realize their health condition early enough. In this study, we propose a data mining based model for early diagnosis and prediction of diabetes using the UCI database. Although K-means is simple and can be used for a wide variety of data types, it is quite sensitive to initial positions of cluster centers which determine the final cluster result, which either provides a sufficient and efficiently clustered dataset for the logistic regression model, or gives a lesser amount of data as a result of incorrect clustering of the original dataset, thereby limiting the performance of the logistic regression model. Our findings offer insights into the comparative strengths and weaknesses of each method, shedding light on their potential applications in diabetes diagnosis and risk assessment. A further experiment with a new dataset showed the applicability of our model for the predication of diabetes.

Pronunciation and Spelling Accuracy in English Words with Initial and Final Consonant Clusters by Arabic-Speaking EFL Learners

Arabic phonotactics significantly differ from English phonotactics in that they usually follow a framework that forbids the presence of consonant clusters in syllabic onsets. This study examines the relationship between Arabic-speaking EFL learners’ spelling accuracy and pronunciation, concentrating on the difficulties caused by consonant clusters in initial and final positions. A mixed-methods approach was employed, involving audio recordings to assess pronunciation accuracy and a spelling test to evaluate spelling performance. According to the results, words with initial consonant clusters (clusters at the beginning of a syllable, or onset) are harder to pronounce and spell than words with final consonant clusters (clusters that follow the vowel). Initial consonant clusters are considered to demand a quick transition in consonants at the start of a syllable, requiring more exact articulation. On the other hand, final clusters allow for a more progressive and straightforward articulation by involving a consonant transition following the initial vowel sound. Additionally, epenthesis may be used as a method to break up consonant clusters in L1 phonological interference, which is consistent with Arabic phonotactic patterns.

Transcriptomic analysis reveals role of lncRNA LOC100257036 to regulate AGAMOUS during cluster compactness of Vitis vinifera cv. sistan yaghooti

Yaghooti grape, as the earliest grape variety in Iran, is considered as more resistant to heat, drought, and salinity than other cultivars. Cluster compactness is regarded as an inappropriate feature for the productivity of Yaghooti grape as a critical commercial and nutritional product. In plants, lncRNAs play a critical role in regulating biological processes related to growth and development. However, the potential role of lncRNAs was not assessed in cluster compactness. Totally, 1549 lncRNAs were identified by RNA-Seq data analysis in three steps of cluster formation, berry formation, and final cluster size after a thorough screening process. In addition, 229 lncRNAs were differentially expressed in the cluster development steps. Based on the functional analysis, lncRNAs are related to AG and MYB, bHLH, LBD, NAC, and WRKY TFs. Further, the target genes enrichment analysis revealed a relationship between lncRNAs with grape growth and development, as well as resistance to abiotic stresses such as heat and drought, plant defense against pathogens, and early grapes ripening. The study identified four lncRNAs as precursors of miRNAs, predicting that 112 other lncRNAs could potentially be targeted by 166 miRNAs. The results provide new insights into the regulatory functions of lncRNAs in Yaghooti grape to improve overall understanding of the molecular mechanisms related to grape compactness.

Interpreting the space characteristics of everyday heritage gardens of Suzhou, China, through a space syntax approach

ABSTRACT China’s everyday heritage gardens, epitomizing social and spatial logic through classical layouts, offer a window into the nexus between Chinese culture and spatial design. Using space syntax and statistical methods, we examined 27 such gardens in Suzhou, categorizing them into four groups by spatial attributes. We first quantified building spatial traits, then delved into their syntactical relationships, guiding the final clustering. Cluster A, with 17 gardens, marries architectural diversity with spatial harmony, ensuring navigability. Cluster B’s seven gardens prioritize open landscapes, minimizing buildings for clarity. Cluster C, comprising two expansive gardens, blends architectural variety with spatial coherence, albeit with some navigational complexity. Unique, Cluster D, solely Yi Garden, presents a rich spatial and architectural tapestry, demanding deeper visitor engagement. This work sheds light on the intertwining of garden design and spatial structures, underscoring influences like ownership and cultural milieu, suggesting broader applicability to heritage gardens.

Assessment of the coherence of groundwater levels in coastal aquifers with climate change and anthropogenic activity

This study aimed to assess the coherence between groundwater levels and various factors during two distinct periods, 2002–2020 and 2025–2050, in Miandoab aquifer in northwestern of the Iran. Partial wavelet coherence and multi-wavelet coherence analyses were employed to assess the coherence between individual parameters and their simultaneous coherence. The factors considered in the study were derived from remote sensing data, including Gravity Recovery and Climate Experiment data and Landsat data, which were utilized to examine water storage anomalies and anthropogenic activity, respectively. Additionally, General Circulation Models were employed to predict groundwater levels under future climate change scenarios via a feedforward neural network. To streamline the modeling process and categorize piezometers, with each group reflecting different patterns, clustering techniques were applied to group multiple piezometers. There were four final clusters, and representative piezometers from each cluster were chosen as targets for modeling and future predictions. Finally, the differences in coherence between past and future periods were compared and analyzed. The results revealed decreasing trends in groundwater level, precipitation and soil moisture index in 2025–2050; however, there were increasing trends in normalized difference vegetation index and temperature. In addition, wavelet analysis indicated that during the period 2025–2050, the delay in interaction between groundwater level and various factors decreased to 0–4 months, whereas longer delays were observed for the period 2002–2020. The analysis of multi- wavelet coherence showed that the combination of climate change and anthropogenic activity may have more significant coherence (0.9–1) with groundwater level than the combination of gravity recovery and climate experiment data and soil moisture index. The results highlight the greater significance of gravity recovery and climate experiment data in terms of coherence with groundwater levels compared to other factors.

A Structured Bipartite Graph Learning method for ensemble clustering

Given a set of base clustering results, conventional bipartite graph-based ensemble clustering methods typically require computing a sample-cluster similarity matrix from each base clustering result. These matrices are then either concatenated or averaged to form a bipartite weight matrix, which is used to create a bipartite graph. Graph-based partition techniques are subsequently applied to this graph to obtain the final clustering result. However, these methods often suffer from unreliable base clustering results, making it challenging to identify a clear cluster structure due to the variations in cluster structures across the base results. In this paper, we propose a novel Structured Bipartite Graph Learning (SBGL) method. Our approach begins by computing a sample-cluster similarity matrix from each base clustering result and constructing a base bipartite graph from each of these matrices. We assume these base bipartite graphs contain a set of latent clusters and project them into a set of sample-latent-cluster bipartite graphs. These new graphs are then ensembled into a bipartite graph with a distinct cluster structure, from which the final set of clusters is derived. Our method allows for different numbers of clusters across base clusterings, leading to improved performance. Experimental results on both synthetic and real-world datasets demonstrate the superior performance of our new method.

Exploring and prioritizing content to include in a medication self-management toolkit for persons with spinal cord injury/dysfunction: A concept mapping approach.

Adults with spinal cord injury/dysfunction (SCI/D) face challenges with medications they take to manage their secondary conditions (e.g., pain, urinary tract infections, autonomic dysreflexia). With many healthcare providers typically involved in care, there are additional challenges with care fragmentation and self-management. Prior research emphasized the desire for more support with medication self-management among this population. To explore what content should be included in a medication self-management resource (i.e., toolkit) for adults with SCI/D, as well as considerations for delivery from the perspectives of adults with SCI/D, caregivers, healthcare providers, and representatives from community organizations. A concept mapping study was conducted. Participants took part in one or more of three activities: brainstorming; sorting and rating; and mapping. Participants generated ideas about the content to include in a medication self-management toolkit. Participants sorted the statements into conceptual piles and assigned a name to each. All statements were rated on a five-point Likert-type scale on importance and realistic to include in the toolkit. Participants decided on the final cluster map, rearranged statements, and assigned a name to each cluster to create visual representations of the data. Forty-four participants took part in this study. The final map contained eight clusters: 1) information-sharing and communication; 2) healthcare provider interactions and involvement; 3) peer and community connections; 4) supports and services for accessing prescription medications and medication information; 5) information on non-prescription medication and medication supplies; 6) safety and lifestyle considerations; 7) general medication information; and 8) practical information and strategies related to medication-taking. Safety and lifestyle considerations was rated as the most important and realistic to include in the toolkit. Given the limited tools to help adults with SCI/D with managing their medications, there is great potential to better support this population across all areas of medication self-management.

Augmented ELBO regularization for enhanced clustering in variational autoencoders

With significant advances in deep neural networks, various new algorithms have emerged that effectively model latent structures within data, surpassing traditional clustering methods. Each data point is expected to belong to a single cluster in a typical clustering algorithm. However, clustering based on variational autoencoders (VAEs) represents the expectation of the overall clusters, denoted as c=1,…,K in the KL divergence term. Consequently, the latent embedding z can be learned to exist across multiple clusters with relatively balanced probabilities, rather than being strongly associated with a specific cluster. This study introduces an additional regularizer to encourage the latent embedding z to have a strong affiliation with specific clusters. We introduce optimization methods to maximize the ELBO that includes the newly added regularization term and explore methods to eliminate computationally challenging terms. The positive impact of this regularization on clustering accuracy was verified by examining the variance of the final cluster probabilities. Furthermore, an enhancement in the clustering performance was observed when regularization was introduced.

Uncertain data density peak clustering algorithm based on JS divergence

Aiming at the defects of traditional density-based uncertainty clustering algorithms, such as parameter sensitivity and poor clustering results for complex manifold uncertain data sets, a new uncertainty data density peak clustering algorithm based on JS divergence (UDPC-JS) is proposed. The algorithm first removes noise points by using the uncertain natural neighbor density factor defined by the uncertain natural neighbor. Secondly, the local density of uncertain data objects is calculated by combining the uncertain natural neighbor and JS divergence. The initial clustering center of the uncertain data set is found by combining the idea of representative points, and the distance between the initial clustering centers is defined based on JS divergence and graph. Then, the decision graph is constructed on the initial clustering center using the local density calculated based on the uncertain natural neighbor and JS divergence and the newly defined distance between the initial clustering centers based on JS divergence and graph, and the final clustering center is selected according to the decision graph. Finally, the unassigned uncertain data objects are assigned to the cluster where their initial clustering center is located. Experimental results show that the algorithm has better clustering effect and accuracy than the comparison algorithm, and has a greater advantage in processing complex manifold uncertain data sets.

Density clustering method based on k-nearest neighbor propagation

Abstract Density clustering, an effective data analysis tool, performs well on arbitrary shapes and non-convex datasets. However, it still has some limitations in identifying the cluster structures of datasets with irregular shapes and uneven density distribution. Aiming at the above problem, this paper proposes a density clustering method based on k-nearest neighbor propagation. Firstly, based on the theory of k-nearest neighbor, the nearest hub points are defined to replace the sample points within their k-nearest neighbors, and the cluster boundary is ascertained relying on the propagation of the nearest hub points with density gradient trend so as to divide the nearest hub points into multiple groups; secondly, a group merging method of the nearest hub points based on the set similarity is given to obtain the set of nearest hub points; and finally, the remaining sample points are assigned to the set of their nearest hub points to obtain the final clusters. The correctness and effectiveness of the algorithm are verified by experiments on synthetic and UCI datasets, and the results show that the algorithm can better cope with the clustering needs of multiple complex data distributions compared with other comparative algorithms.

Towards attributed graph clustering using enhanced graph and reconstructed graph structure

Attributed graph clustering, leveraging both structural and attribute information, is crucial in various real-world applications. However, current approaches face challenges stemming from the sparsity of graphs and sensitivity to noise in Graph Convolutional Networks (GCNs). Moreover, GCN-based methods are often designed based on the assumption of homophilic graph and ignore heterophilic graph. To address these, we propose a graph clustering method that consists of four phases: graph enhance, graph reconstruction, graph refine, and dual-guidance supervisor module. An enhanced graph module is defined by an auxiliary graph to consider distant relationships in the topology structure to alleviate the limitations of sparse graphs. The graph reconstruction phase includes the creation and integration of homophily and heterophily graphs to achieve graph-agnostic. In graph refine, the auxiliary graph is iteratively improved to enhance the generalization of the representation. In this phase, a subspace clustering module is applied to convert attribute-based embeddings into relationship-based representations. Finally, the extracted graphs are fed to a dual-guidance supervisor module to obtain the final clustering result. Experimental validation on several benchmark datasets demonstrates the efficiency of our model. Meanwhile, the findings offer significant advancements in attributed graph clustering, promising improved applicability in various domains.

Multi-view attributed graph clustering based on graph diffusion convolution with adaptive fusion

Graph clustering is one of the popular techniques in data mining, which is widely used in social network analysis, recommendation system and anomaly detection. The existing methods based on multi-view attributed graph ignore the difference of different views and the consistency between views. To this end, this paper proposes a novel method for multi-view attributed graph clustering based on graph diffusion convolution with adaptive fusion. Specifically, a graph diffusion convolution is firstly employed to propagate features of the adjacency and attribute matrices on each view. A linear encoder–decoder is then used to learn the graph embedding representation of each view. Meanwhile, the consistency learning module is applied to capture the geometric relationship consistency of different views for finding a consistent embedding space during the embedding representation learning. To measure the importance of different views, the attention mechanism is subsequently utilized to adaptively learn the weights of different views to fuse multiple views of data for the final clustering. Finally, the graph representation learning and clustering tasks are jointly optimized to obtain a clustering friendly graph representation. Experiments conducted on two public datasets and three real futures datasets from the Zhengzhou Commodity Exchange demonstrate the effectiveness of the proposed method in the task of graph clustering.

The Improvement of Density Peaks Clustering Algorithm and Its Application to Point Cloud Segmentation of LiDAR.

This work focuses on the improvement of the density peaks clustering (DPC) algorithm and its application to point cloud segmentation in LiDAR. The improvement of DPC focuses on avoiding the manual determination of the cut-off distance and the manual selection of cluster centers. And the clustering process of the improved DPC is automatic without manual intervention. The cut-off distance is avoided by forming a voxel structure and using the number of points in the voxel as the local density of the voxel. The automatic selection of cluster centers is realized by selecting the voxels whose gamma values are greater than the gamma value of the inflection point of the fitted γ curve as cluster centers. Finally, a new merging strategy is introduced to overcome the over-segmentation problem and obtain the final clustering result. To verify the effectiveness of the improved DPC, experiments on point cloud segmentation of LiDAR under different scenes were conducted. The basic DPC, K-means, and DBSCAN were introduced for comparison. The experimental results showed that the improved DPC is effective and its application to point cloud segmentation of LiDAR is successful. Compared with the basic DPC, K-means, the improved DPC has better clustering accuracy. And, compared with DBSCAN, the improved DPC has comparable or slightly better clustering accuracy without nontrivial parameters.

γ-clustering problems: Classical and parametrized complexity

We introduce the γ-clustering problems, which are variants of the well-known Cluster Editing/Deletion/Completion problems, and defined as: given a graph G, how many edges must be edited in G, deleted from G, or added to G in order to have a disjoint union of γ-quasi-cliques. We provide here the complete complexity classification of these problems along with FPT algorithms parameterized by the number of modifications, for the NP-complete problems. We also study here a variant of these problems where the number of final clusters is a fixed constant, obtaining mostly the same results regarding classical and parameterized complexity.

Anchor-based fast spectral ensemble clustering

Ensemble clustering can obtain better and more robust results by fusing multiple base clusterings, which has received extensive attention. Although many representative algorithms have emerged in recent years, this field still has two tricky problems. First, spectral clustering can identify clusters of arbitrary shapes, but the high time and space complexity limit its application in generating base clusterings. Most existing algorithms utilize k-means to generate base clusterings, and the clustering effect on nonlinearly separable datasets needs further improvement. Second, ensemble clustering algorithms should generate multiple base clusterings. Even if low-complexity algorithms are applied, the running time is also long, which seriously affects the application of ensemble clustering algorithms on large-scale datasets. To tackle these problems, we propose a fast K-nearest neighbors approximation method, construct an anchor graph to approximate the similarity matrix, and use singular value decomposition (SVD) instead of eigenvalue decomposition (EVD) to reduce the time and space complexity of conventional spectral clustering. At the same time, we obtain multiple base clusterings by running spectral embedding once. Finally, we convert these base clusterings into a bipartite graph and use transfer cut to get the final clustering results. The proposed algorithms significantly reduce the running time of ensemble clustering. Experimental results on large-scale datasets fully prove the efficiency and superiority of our proposed algorithm.

Flexible density peak clustering for real-world data

In density based clustering, the density peak algorithm has attracted much attention due to its effectiveness and simplicity, and a vast amount of clustering approaches have been proposed based on this algorithm. Some of these works require manual selection of cluster centers with a decision graph, where human involvement leads to uncertainty in clustering results. In order to avoid human involvement, some other algorithms depend on user-specified number of clusters to determine cluster centers automatically. However, it is well known that accurate estimation of number of clusters is a long-standing difficulty in data clustering. In this paper we present a sequential density peak clustering algorithm to extract clusters one by one, thereby determining the number of clusters automatically and avoiding manual selection of cluster centers in the meanwhile. Starting from a density peak, our algorithm generates an initial cluster surrounding the density peak in the first step, and then obtains the final cluster by expanding the initial cluster based on the relative density relationship among neighboring data points. With a peeling-off strategy, we obtain all the clusters sequentially. Our algorithm works well with clusters of Gaussian distribution and is therefore potential for clustering of real-world data. Experiments with a large number of synthetic and real datasets and comparisons with existing algorithms demonstrate the effectiveness of the proposed algorithm.

Semi-supervised clustering guided by pairwise constraints and local density structures

Clustering based on local density peaks and graph cut (LDP-SC) is one of the state-of-the-art algorithms in unsupervised clustering, which first divides the data set to be multiple local trees, and then aggregates these local trees to obtain the final clustering result. However, for complex data sets, there might exist data points from different classes in the same local tree. In this article, we use pairwise constraint information to resolve this issue and propose a semi-supervised local density peaks and graph cut based clustering algorithm (SLDPC). In particular, SLDPC proposes intra-cluster conflict resolution and inter-cluster conflict resolution steps to split the local trees which are inconsistent with the provided pairwise constraint information. Theoretically, we show that the two steps will finish in a finite number of operations and the split local trees will be consistent with the pairwise constraint information. Subsequently, root node redirection and noise filtering steps are designed to avoid the local trees becoming too fragmented. Finally, we exploit the E2CP algorithm to further improve the similarity matrix between local trees using the pairwise constraint information, and the spectral clustering algorithm is adopted to obtain the clustering result. Experiments on multiple widely used synthetic and real-world data sets show that SLDPC is superior to LDP-SC and several other semi-supervised prominent clustering algorithms for most of the cases.

Re-Purposing the Ordering of Routine Laboratory Tests in Hospitalized Medical Patients (RePORT): protocol for a multicenter stepped-wedge cluster randomised trial to evaluate the impact of a multicomponent intervention bundle to reduce laboratory test over-utilization

BackgroundLaboratory test overuse in hospitals is a form of healthcare waste that also harms patients. Developing and evaluating interventions to reduce this form of healthcare waste is critical. We detail the protocol for our study which aims to implement and evaluate the impact of an evidence-based, multicomponent intervention bundle on repetitive use of routine laboratory testing in hospitalized medical patients across adult hospitals in the province of British Columbia, Canada.MethodsWe have designed a stepped-wedge cluster randomized trial to assess the impact of a multicomponent intervention bundle across 16 hospitals in the province of British Columbia in Canada. We will use the Knowledge to Action cycle to guide implementation and the RE-AIM framework to guide evaluation of the intervention bundle. The primary outcome will be the number of routine laboratory tests ordered per patient-day in the intervention versus control periods. Secondary outcome measures will assess implementation fidelity, number of all common laboratory tests used, impact on healthcare costs, and safety outcomes. The study will include patients admitted to adult medical wards (internal medicine or family medicine) and healthcare providers working in these wards within the participating hospitals. After a baseline period of 24 weeks, we will conduct a 16-week pilot at one hospital site. A new cluster (containing approximately 2–3 hospitals) will receive the intervention every 12 weeks. We will evaluate the sustainability of implementation at 24 weeks post implementation of the final cluster. Using intention to treat, we will use generalized linear mixed models for analysis to evaluate the impact of the intervention on outcomes.DiscussionThe study builds upon a multicomponent intervention bundle that has previously demonstrated effectiveness. The elements of the intervention bundle are easily adaptable to other settings, facilitating future adoption in wider contexts. The study outputs are expected to have a positive impact as they will reduce usage of repetitive laboratory tests and provide empirically supported measures and tools for accomplishing this work.Trial RegistrationThis study was prospectively registered on April 8, 2024, via ClinicalTrials.gov Protocols Registration and Results System (NCT06359587). https://classic.clinicaltrials.gov/ct2/show/NCT06359587?term=NCT06359587&recrs=ab&draw=2&rank=1