Clustering Accuracy Research Articles

Enhancing Clustering Stability and Efficiency: A Framework for Optimizing K-means, K-medoids, and K-shape with Intelligent Algorithms

Clustering methods like Kmeans often produce inconsistent results due to the random initialization of cluster centroids, even with optimizations such as Kmeans++, which improve centroid selection but fail to eliminate sensitivity to initialization randomness. Additionally, the choice of the number of clusters and distance metrics significantly impacts clustering performance. This paper proposes an enhanced framework combining intelligent optimization algorithms—Sparrow Search Algorithm (SSA), Dung Beetle Optimizer (DBO), and Sine Cosine Algorithm (SCA)—to optimize clustering outcomes for Kmeans, Kmedoids, and Kshape. The framework also incorporates dimensionality reduction techniques, including Principal Component Analysis (PCA), Non-Negative Matrix Factorization (NNMF), and Singular Value Decomposition (SVD), to address high-dimensional data challenges. Experimental results on benchmark datasets demonstrate the proposed framework’s effectiveness, with SSA achieving the highest silhouette score of 0.68 and reducing runtime by 35% compared to traditional methods. This approach enhances clustering stability and accuracy, offering a robust solution for diverse applications.

The impact of neglecting feature scaling in k-means clustering.

Despite the popularity of k-means clustering, feature scaling before applying it can be an essential yet often neglected step. In this study, feature scaling via five methods: Z-score, Min-Max normalization, Percentile transformation, Maximum absolute scaling, or RobustScaler beforehand was compared with using the raw (i.e., non-scaled) data to analyze datasets having features with different or the same units via k-means clustering. The results of an experimental study show that, for features with different units, scaling them before k-means clustering provided better accuracy, precision, recall, and F-score values than when using the raw data. Meanwhile, when features in the dataset had the same unit, scaling them beforehand provided similar results to using the raw data. Thus, scaling the features beforehand is a very important step for datasets with different units, which improves the clustering results and accuracy. Of the five feature-scaling methods used in the dataset with different units, Z-score standardization and Percentile transformation provided similar performances that were superior to the other or using the raw data. While Maximum absolute scaling, slightly more performances than the other scaling methods and raw data when the dataset contains features with the same unit, the improvement was not significant.

Nonlinear subspace clustering by functional link neural networks

Nonlinear subspace clustering based on a feed-forward neural network has been demonstrated to provide better clustering accuracy than some advanced subspace clustering algorithms. While this approach demonstrates impressive outcomes, it involves a balance between effectiveness and computational cost. In this study, we employ a functional link neural network to transform data samples into a nonlinear domain. Subsequently, we acquire a self-representation matrix through a learning mechanism that builds upon the mapped samples. As the functional link neural network is a single-layer neural network, our proposed method achieves high computational efficiency while ensuring desirable clustering performance. By incorporating the local similarity regularization to enhance the grouping effect, our proposed method further improves the quality of the clustering results. We name our method as Functional Link Neural Network Subspace Clustering (FLNNSC). Furthermore, we propose a convex combination subspace clustering scheme that combines a linear subspace clustering method with the functional link neural network subspace clustering approach. This combination method is named as Convex Combination Subspace Clustering (CCSC), which allows for a dynamic balance between linear and nonlinear representations. Extensive experiments conducted on four widely used datasets, including Extended Yale B, USPS, COIL20, and ORL, demonstrate that both FLNNSC and CCSC outperform several state-of-art subspace clustering methods in terms of clustering accuracy. Our affinity graph experiments reveal that FLNNSC exhibits clear block diagonal structures. We provide recommendations for hyperparameters in FLNNSC by performing a parameter sensitivity analysis, and empirically verify the convergence of FLNNSC. Additionally, we show that FLNNSC has a lower computational cost compared to two high-performing methods.

Analyzing the Accuracy of Clustering Based Cab Recommender System (CBCRS) Across Varying Densities

Analyzing the Accuracy of Clustering Based Cab Recommender System (CBCRS) Across Varying Densities

Joint sparse subspace clustering via fast [formula omitted]-norm constrained optimization

Subspace clustering gains popularity in unsupervised machine learning due to its excellent dimensionality reduction capability and interpretability. Although existing research has made significant progress in clustering performance, it still faces two limitations. The first is the challenge of poor feature representation, which makes it hard to explore global structural sparsity. The second is the issue of long runtime, which makes it computationally inefficient. In this paper, we propose a joint sparse subspace clustering (JSSC) method to capture the structural features in the representation matrix via the ℓ2,0-norm constraint, which encourages sparsity across rows and leads to enhance clustering performance. In algorithms, an optimization scheme based on proximal alternating minimization (PAM) is developed, in which each subproblem has a closed-form solution, thus ensuring the effectiveness. Moreover, the convergence guarantee is rigorously proved in theory. Numerical experiments on image and hyperspectral datasets validate its excellent performance in terms of high accuracy with a high computational speed. In particular, the proposed method improves the clustering accuracy by at least 6.1% and the runtime by at least 14 times compared with competitors on the ORL dataset. The code of our proposed JSSC is available at https://github.com/zhudafa/JSSC.

Ultraweak light-modulated heterostructure with bidirectional photoresponse for static and dynamic image perception

The human visual system’s adaptability to varying brightness levels has inspired the development of optoelectronic neuromorphic devices. However, achieving bidirectional photoresponse, essential for mimicking these functions, often requires high operation voltages or high light intensities. Here, we propose a bidirectional ZnO/CsPbBr3 heterostructure based neuromorphic image sensor array (10 × 10 pixels) capable of ultraweak light stimulation. The device demonstrates positive and negative photoconductivity through the ionization and deionization of oxygen vacancies in the ZnO channel, extendable to other ZnO/perovskites and IGZO/perovskites heterostructures. Operating at a reduced bias voltage of 2.0 V, the array achieves synaptic weight updates under green (525 nm) and UV (365 nm) light with light intensities ranging from as low as 45 nW/cm² to 15.69 mW/cm², mimicking basic synaptic functions and visual adaptation. It performs multiple image pre-processing tasks, including background denoising and encoding spatiotemporal motion, achieving 92% accuracy in pattern recognition and 100% accuracy in motion clustering. This straightforward strategy highlights a potential for intelligent visual systems capable of real-time image processing under low voltage and dark conditions.

ScVAG: Unified single-cell clustering via variational-autoencoder integration with Graph Attention Autoencoder

Single-cell RNA sequencing (scRNA-seq) enables high-resolution transcriptional profiling of cell heterogeneity. However, analyzing this noisy, high-dimensional matrix remains challenging. We present scVAG, an integrated deep learning framework combining Variational-Autoencoder (VAE) and Graph Attention Autoencoder (GATE) for enhanced single-cell clustering. Building upon scGAC, our approach replaces its restrictive linear principal component analysis (PCA) with nonlinear dimensionality reduction better suited for scRNA-seq data. Specifically, we integrate VAE and GATE to enable more flexible latent space encoding. Extensive experiments on 20 datasets demonstrate scVAG's superior performance over previous state-of-the-art methods including scGAC, SCEA, SC3, Seurat, scGNN, scASGC, DESC, NIC, scLDS2, DRJCC, sLMIC, and jSRC. On average, scVAG improves clustering accuracy by 5 percent in ARI and 4 percent in NMI parameters. Visualizations highlight scVAG's capacity to recover interpretable biological structures. Our VAE-GATE pipeline extracts intricate expression patterns into compact representations that precisely delineate cell subpopulations consistent with ground truth labels. Overall, scVAG establishes a robust architecture for elucidating cell taxonomies from noisy transcriptomic inputs.

Application of big data and artificial intelligence in visual communication art design.

In the era of continuous development of computer technology, the application of artificial intelligence (AI) and big data is becoming more and more extensive. With the help of powerful computer and network technology, the art of visual communication (VISCOM) has ushered in a new chapter of digitalization and intelligence. How vision can better perform interdisciplinary and interdisciplinary artistic expression between art and technology and how to use more novel technology, richer forms, and more appropriate ways to express art has become a new problem in visual art creation. This essay aims to investigate and apply VISCOM art through big data and AI methods. This essay proposed the STING algorithm for big data for multi-resolution information clustering in VISCOM art. In addition, the convolutional neural network (CNN) in AI technology was used to identify the conveyed objects or scenes to achieve the purpose of designing art with different characteristics for different scenes and groups of people. STING is a multi-resolution clustering technique for big data, with the advantage of efficient data processing. In the experimental part, this essay selected a variety of design contents in VISCOM art, including logo design, text design, scene design, packaging design and poster design. STING and CNN algorithms were used to cluster and AI-identify the design elements 16 of the design projects might contain. The results showed that the overall average clustering accuracy was above 82%, the accuracy of scene element recognition mainly was above 80%, and the accuracy of facial recognition was above 80%; this showed that this essay applied AI and big data to the design of VISCOM, and had a good effect on the clustering and identification of design elements. According to expert scores, these applications' reliability and practicality scores were above 70 points, with an average of about 80 points. Therefore, applying big data and AI to VISCOM in this essay is reliable and feasible.

PartIES: a disease subtyping framework with Partition-level Integration using diffusion-Enhanced Similarities from multi-omics Data.

Integrating multi-omics data helps identify disease subtypes. Many similarity-based methods were developed for disease subtyping using multi-omics data, with many of them focusing on extracting common clustering structures across multiple types of omics data, but not preserving data-type-specific clustering structures. Moreover, clustering performance of similarity-based methods is affected when similarity measures are noisy. Here we proposed PartIES, a Partition-level Integration using diffusion-Enhanced Similarities to perform disease subtyping using multi-omics data. PartIES uses diffusion to reduce noises in individual similarity/kernel matrices from individual omics data types first, and then extract partition information from diffusion-enhanced similarity matrices and integrate the partition-level similarity through a weighted average iteratively. Simulation studies showed that (1) the diffusion step enhances clustering accuracy, and (2) PartIES outperforms competing methods, particularly when omics data types provide different clustering structures. Using mRNA, long noncoding RNAs, microRNAs expression data, DNA methylation data, and somatic mutation data from The Cancer Genome Atlas project, PartIES identified subtypes in bladder urothelial carcinoma, liver hepatocellular carcinoma, and thyroid carcinoma that are most significantly associated with patient survival across all methods. Further investigations suggested that among subtype-associated genes, many of those that are highly interacting with other genes are known important cancer genes. The identified cancer subtypes also have different activity levels for some known cancer-related pathways. The R code can be accessed at https://github.com/yuqimiao/PartIES.git.

Dimensionality Reduction for Data Analysis With Quantum Feature Learning

ABSTRACTTo improve data analysis and feature learning, this study compares the effectiveness of quantum dimensionality reduction (qDR) techniques to classical ones. In this study, we investigate several qDR techniques on a variety of datasets such as quantum Gaussian distribution adaptation (qGDA), quantum principal component analysis (qPCA), quantum linear discriminant analysis (qLDA), and quantum t‐SNE (qt‐SNE). The Olivetti Faces, Wine, Breast Cancer, Digits, and Iris are among the datasets used in this investigation. Through comparison evaluations against well‐established classical approaches, such as classical PCA (cPCA), classical LDA (cLDA), and classical GDA (cGDA), and using well‐established metrics like loss, fidelity, and processing time, the effectiveness of these techniques is assessed. The findings show that cPCA produced positive results with the lowest loss and highest fidelity when used on the Iris dataset. On the other hand, quantum uniform manifold approximation and projection (qUMAP) performs well and shows strong fidelity when tested against the Wine dataset, but ct‐SNE shows mediocre performance against the Digits dataset. Isomap and locally linear embedding (LLE) function differently depending on the dataset. Notably, LLE showed the largest loss and lowest fidelity on the Olivetti Faces dataset. The hypothesis testing findings showed that the qDR strategies did not significantly outperform the classical techniques in terms of maintaining pertinent information from quantum datasets. More specifically, the outcomes of paired t‐tests show that when it comes to the ability to capture complex patterns, there are no statistically significant differences between the cPCA and qPCA, the cLDA and qLDA, and the cGDA and qGDA. According to the findings of the assessments of mutual information (MI) and clustering accuracy, qPCA may be able to recognize patterns more clearly than standardized cPCA. Nevertheless, there is no discernible improvement between the qLDA and qGDA approaches and their classical counterparts.

An Entropy-Based Clustering Algorithm for Real-Time High-Dimensional IoT Data Streams.

The rapid growth of data streams, propelled by the proliferation of sensors and Internet of Things (IoT) devices, presents significant challenges for real-time clustering of high-dimensional data. Traditional clustering algorithms struggle with high dimensionality, memory and time constraints, and adapting to dynamically evolving data. Existing dimensionality reduction methods often neglect feature ranking, leading to suboptimal clustering performance. To address these issues, we introduce E-Stream, a novel entropy-based clustering algorithm for high-dimensional data streams. E-Stream performs real-time feature ranking based on entropy within a sliding time window to identify the most informative features, which are then utilized with the DenStream algorithm for efficient clustering. We evaluated E-Stream using the NSL-KDD dataset, comparing it against DenStream, CluStream, and MR-Stream. The evaluation metrics included the average F-Measure, Jaccard Index, Fowlkes-Mallows Index, Purity, and Rand Index. The results show that E-Stream outperformed the baseline algorithms in both clustering accuracy and computational efficiency while effectively reducing dimensionality. E-Stream also demonstrated significantly less memory consumption and fewer computational requirements, highlighting its suitability for real-time processing of high-dimensional data streams. Despite its strengths, E-Stream requires manual parameter adjustment and assumes a consistent number of active features, which may limit its adaptability to diverse datasets. Future work will focus on developing a fully autonomous, parameter-free version of the algorithm, incorporating mechanisms to handle missing features and improving the management of evolving clusters to enhance robustness and adaptability in dynamic IoT environments.

Large data density peak clustering based on sparse auto-encoder and data space meshing via evidence probability distribution

The development of big data analysis technology has brought new development opportunities to the production and management of various industries. Through the mining and analysis of various data in the operation process of enterprises by big data technology, the internal associated data of the enterprises and even the entire industry can be obtained. As a common method for large-scale data statistical analysis, clustering technology can effectively mine the relationship within massive heterogeneous multidimensional data, complete unlabeled data classification, and provide data support for various model analysis of big data. Common big data density clustering methods are time-consuming and easy to cause errors in data density allocation, which affects the accuracy of data clustering. Therefore we propose a novel large data density peak clustering based on sparse auto-encoder and data space meshing via evidence probability distribution. Firstly, the sparse auto-encoder in deep learning is used to achieve feature extraction and dimensionality reduction for input high-dimensional data matrix through training. Secondly, the data space is meshed to reduce the calculation of the distance between the sample data points. When calculating the local density, not only the density value of the grid itself, but also the density value of the nearest neighbors are considered, which reduces the influence of the subjective selection truncation distance on the clustering results and improves the clustering accuracy. The grid density threshold is set to ensure the stability of the clustering results. Using the K-nearest neighbor information of the sample points, the transfer probability distribution strategy and evidence probability distribution strategy are proposed to optimize the distribution of the remaining sample points, so as to avoid the joint error of distribution. The experimental results show that the proposed algorithm has higher clustering accuracy and better clustering performance than other advanced clustering algorithms on artificial and real data sets.

Two Fuzzy Clustering Algorithms Based on ARMA Model

This study proposes two fuzzy clustering algorithms based on autoregressive moving average (ARMA) model for series data. The first, referred to as Tsallis entropy-regularized fuzzy c-ARMA model (TFCARMA), is created from k-ARMA, a conventional hard clustering algorithm for series data. TFCARMA is motivated by the relationship between the two clustering algorithms for vectorial data: k-means and Tsallis entropy-regularized fuzzy c-means. The second, referred to as q-divergence-based fuzzy c-ARMA model (QFCARMA), is created from ARMA mixtures, a conventional probabilistic clustering algorithm for series data. QFCARMA is motivated by the relationship between the two clustering algorithms for vectorial data: Gaussian mixture model and q-divergence-based fuzzy c-means. Based on numerical experiments using an artificial dataset, we observed the effects of fuzzification parameters in the proposed algorithms and relationship between the proposed and conventional algorithms. Moreover, numerical experiments using seven real datasets compared the clustering accuracy among the proposed and conventional algorithms.

Random Forest-Based Forensic Investigation of Non-Fungible Tokens: for Enhanced Detection and Anomaly Identification

The proposed work builds upon the Random Forest machine learning algorithm to improve the process of digit forensic investigation in case of NFT. The following structure of this framework is aimed at identifying and disabling fraudulent or suspicious activities in NFT transactions by comparing different parameters like the Detection Time, False Positive Rate, the Total Transaction Volume Analyzed, the Anomalous Transaction Ratio, Clustering Accuracy, Data Utilization Efficiency, and Detection Sensitivity. Through using Random Forest, a solid ensemble learning technique that is well known for its on high accuracy as well as off overfitting tendency, it optimistically improves the identifying abilities of the framework in isolation of the false positives. The ability of the proposed system to deliver optimal results is further explained by line plots, area charts, histograms, and stem plots which all provide the variation of these metrics as the time proceeds. Not only does it enhance the effectiveness of detecting the fraudulent transactions, but it also enhances the application of data in the forensic analysis that creates a great advantage in the increasing realm of digital assets for investigators.

Unsupervised person re‐identification based on adaptive information supplementation and foreground enhancement

AbstractUnsupervised person re‐identification has attracted vital interest because of its ability to protect privacy, significantly lower the expense of manual annotation, and eliminate the need for data labels. General unsupervised methods train the network only through global features, which causes the fine‐grained information contained in local features to be ignored in the recognition process, resulting in large amounts of label noise and affecting the recognition accuracy. Moreover, more robust pedestrian features can also improve the accuracy of clustering and enable unsupervised person re‐identification to obtain better results. To address these issues, first, a dual‐branch structure was proposed, which separately obtains the global features of the pedestrian and the local features by dividing the global features into a few equal sections. Then, an adaptive information supplementation (AIS) method based on the k‐nearest neighbor algorithm is designed to ascertain each local feature's relevance to the global features, calculating adaptive weight scores for information supplementation. Finally, these weight scores are used to reallocate the weights of the global features in each part, acquiring features that contain more pedestrian information during the representation learning process. These better features are used to reduce label noise to obtain more accurate pseudo‐labels. Second, an adaptive foreground enhancement module (AFEM) was proposed and inserted before clustering to increase the robustness of pedestrian features, which increases the precision of the pseudo‐labels that are produced after clustering. Experiments on Market‐1501, DukeMTMC‐reID, and MSMT17 demonstrate that the proposed method achieves better results than state‐of‐the‐art methods in fully unsupervised person re‐identification tasks.

Dual-dimensional contrastive learning for incomplete multi-view clustering

Incomplete multi-view clustering (IMVC) is a critical task in real-world applications, where missing data in some views can severely limit the ability to leverage complementary information across views. This issue leads to incomplete sample representations, hindering model performance. Current contrastive learning methods for IMVC exacerbate the problem by directly constructing data pairs from incomplete samples, ignoring essential information and resulting in class collisions, where samples from different classes are incorrectly grouped together due to a lack of label guidance. These challenges are particularly detrimental in fields like recommendation systems and bioinformatics, where accurate clustering of high-dimensional and incomplete data is essential for decision-making. To address these issues, we propose Dual-dimensional Contrastive Learning (DCL), an online IMVC model that fills missing values through multi-view consistency transfer, enabling simultaneous clustering and representation learning via instance-level and cluster-level contrastive learning in both row and column spaces. DCL mitigates class collision issues by generating high-confidence pseudo-labels and using an optimal transport matrix, significantly improving clustering accuracy. Extensive experiments demonstrate that DCL achieves state-of-the-art results across five datasets. The code is available at https://github.com/2251821381/DCL.

Multi-layer multi-level comprehensive learning for deep multi-view clustering

Multi-view clustering has attracted widespread attention because of its capability to identify the common semantics shared by the data captured from different views of data, objects or phenomena. This is a challenging problem but with the emergence of deep auto-encoder networks, the performance of multi-view clustering methods has considerably improved. However, it is notable that most existing methods merely utilize the features outputted by the last encoder layer to carry out the clustering task. Such approach neglects potentially useful information conveyed by the features of the previous layers. To address the this problem, we propose a novel multi-layer multi-level comprehensive learning framework for deep multi-view clustering (3MC). 3MC firstly conducts a contrastive learning involving different views based on deep features in each encoder layer separately, so as to achieve multi-view feature consistency. The next step is to construct layer-specific label MLPs to transform the features in each layer to high-level semantic labels. Finally, 3MC conducts an inter-layer contrastive learning using the high-level semantic labels in order to obtain multi-layer consistent clustering assignments. We demonstrate that the proposed comprehensive learning strategy, commencing from layer specific inter-view feature comparison to inter-layer high-level label comparison extracts and utilizes the underlying multi-view complementary information very successfully and achieves more accurate clustering. An extensive experimental comparison with the state-of-the-art methods demonstrates the effectiveness of the proposed framework. The code of this paper is available at https://github.com/chenzhe207/3MC.

Dimensionality reduction distills complex evolutionary relationships in seasonal influenza and SARS-CoV-2.

Public health researchers and practitioners commonly infer phylogenies from viral genome sequences to understand transmission dynamics and identify clusters of genetically-related samples. However, viruses that reassort or recombine violate phylogenetic assumptions and require more sophisticated methods. Even when phylogenies are appropriate, they can be unnecessary or difficult to interpret without specialty knowledge. For example, pairwise distances between sequences can be enough to identify clusters of related samples or assign new samples to existing phylogenetic clusters. In this work, we tested whether dimensionality reduction methods could capture known genetic groups within two human pathogenic viruses that cause substantial human morbidity and mortality and frequently reassort or recombine, respectively: seasonal influenza A/H3N2 and SARS-CoV-2. We applied principal component analysis, multidimensional scaling (MDS), t-distributed stochastic neighbor embedding (t-SNE), and uniform manifold approximation and projection to sequences with well-defined phylogenetic clades and either reassortment (H3N2) or recombination (SARS-CoV-2). For each low-dimensional embedding of sequences, we calculated the correlation between pairwise genetic and Euclidean distances in the embedding and applied a hierarchical clustering method to identify clusters in the embedding. We measured the accuracy of clusters compared to previously defined phylogenetic clades, reassortment clusters, or recombinant lineages. We found that MDS embeddings accurately represented pairwise genetic distances including the intermediate placement of recombinant SARS-CoV-2 lineages between parental lineages. Clusters from t-SNE embeddings accurately recapitulated known phylogenetic clades, H3N2 reassortment groups, and SARS-CoV-2 recombinant lineages. We show that simple statistical methods without a biological model can accurately represent known genetic relationships for relevant human pathogenic viruses. Our open source implementation of these methods for analysis of viral genome sequences can be easily applied when phylogenetic methods are either unnecessary or inappropriate.

Integrated STL-DBSCAN algorithm for online hydrological and water quality monitoring data cleaning

Online hydrological and water quality monitoring data has become increasingly crucial for water environment management such as assessment and modeling. However, online monitoring data often contains erroneous or incomplete datasets, consequently affecting its operational use. In the study, we developed an automated data cleaning algorithm grounded in Seasonal-Trend decomposition using Loess (STL) and Density-Based Spatial Clustering of Applications with Noise (DBSCAN). STL identifies and corrects more obvious anomalies in the time series, followed by DBSCAN for further refinement, in which the reverse nearest neighbor method was employed to enhance the clustering accuracy. To improve anomaly detection, a two-level residual judgment threshold was applied. The algorithm has been successfully applied to three reservoirs in Shanghai, China, achieving the precision rate of 0.91 and recall rate of 0.81 for dissolved oxygen and pH. The proposed algorithm can be potentially applied for cleaning of environment monitoring data with high accuracy and stability.

Handling Noise Data with PCA Method and Optimization Using Hybrid Fuzzy C-Means and Genetic Algorithm

The significance of machine learning (ML) and data mining techniques particularly clustering is examined in this research, in managing large data sets for customer segmentation in the retail sector. The research emphasizes the challenges posed by data noise and proposes a solution using Principal Component Analysis (PCA) to improve accuracy. This study introduces a hybrid approach that combines Fuzzy C-Means (FCM) with genetic algorithms for optimization in customer segmentation, and suggests further research on the optimal number of clusters and data noise elimination. By addressing data noise, the proposed PCA-based method achieved a higher accuracy rate of 98% compared to 93% without PCA. This finding underscores the effectiveness of PCA in noise reduction, improving clustering accuracy. This research contributes to the advancement of customer-focused business strategies through better data analysis and interpretation. The proposed approach has potential applications in areas including data analysis, pattern recognition, and image processing, highlighting its relevance in the contemporary business environment.