Traditional Clustering Research Articles

Sparse multi-view image clustering with complete similarity information

Multi-view image clustering aims to efficiently divide the collection of images by studying the characteristics of different views. Many studies performed Laplacian dimensionality reduction on the original image to avoid noise interference in constructing the similarity matrix. However, they ignored the problem that local similarity information and isolated image information are lost due to dimensionality reduction. These problems will result in the similarity matrix being insufficient to accurately describe the similarity between images, which in turn affects the clustering accuracy. Therefore, we propose a sparse multi-view spectral clustering model with complete similarity information between images (SMSC-CSI). The model combines the original image space and the low-dimensional spectral embedding space based on the adaptive neighbors method to learn the initial similarity matrices of each view jointly. On the one hand, the original space can retain all the similarity information between images so that the initial similarity matrix can accurately describe the similarity between images, which is conducive to accurate clustering. On the other hand, the low-dimensional space can avoid noise interference and retain the main structure of high-dimensional data, improving the robustness of the initial similarity matrix. Meanwhile, to ensure consistency among the views, the model minimizes the difference between the initial similarity matrix and the central fusion matrix by alternately updating to obtain the optimal weights and central fusion matrix for each view. Finally, we can obtain ideal clustering results directly with low model complexity and without post-processing steps such as K-means by adding non-negative Laplace rank constraints and ℓ0-norm constraints to the model objective function. Experimental results on different real image datasets show that SMSC-CSI can outperform some traditional clustering models and recent models on multi-view image clustering.

Automated lithofluid and facies classification in well logs: The rock-physics perspective

Although an accurate lithofluid and facies interpretation from wireline log data is critical for applications such as joint facies and impedance inversion of seismic data, extracting this information manually is challenging due to the logs’ complexity and the need to combine information from different logs. Traditional clustering methods also struggle with lithofluid type inference due to differing depth trends in petrophysical rock properties stemming from compaction and diagenesis. We introduce a rock-physics machine-learning workflow that automates lithofluid classification and property depth trend modeling to address these challenges. This workflow uses a maximum-likelihood approach, explicitly accounting for depth-related effects via rock-physics models (RPMs) to infer lithofluid types from borehole data. It uses a robust expectation-maximization algorithm to associate each lithofluid type with a specific RPM instance constrained within physically reasonable bounds. The workflow directly outputs lithofluid type proportions and type-specific RPMs with associated uncertainties, providing essential prior information for seismic inversion.

A New Approach In Metaheuristic Clustering: Coot Clustering

As a result of technological advancements, the increase in vast amounts of data in today's world has made artificial intelligence and data mining significantly crucial. In this context, the clustering process, which aims to explore hidden patterns and meaningful relationships within complex datasets by grouping similar features to conduct more effective analyses, holds vital importance. As an alternative to classical clustering methods that face challenges such as large volumes of data and computational complexities, a metaheuristic clustering method utilizing Coot Optimization (COOT), a swarm intelligence-based algorithm, has been proposed. COOT, inspired by the hunting stages of eagles and recently introduced into the literature, is a metaheuristic method. Through the proposed COOT metaheuristic clustering method, the aim is to contribute to the literature by leveraging COOT's robust exploration and exploitation processes, utilizing its dynamic and flexible structure. Comprehensive experimental clustering studies were conducted to evaluate the consistency and effectiveness of the COOT-based algorithm using randomly generated synthetic data and the widely used Iris dataset in the literature. The same datasets underwent analysis using the traditional clustering algorithm K-Means, renowned for its simplicity and computational speed, for comparative purposes. The performance of the algorithms was assessed using cluster validity measures such as Silhouette Global, Davies-Bouldin, Krznowski-Lai, and Calinski-Harabasz indices, along with the Total Squared Error (SSE) objective function. Experimental results indicate that the proposed algorithm performs clustering at a competitive level with K-Means and shows potential, especially in multidimensional datasets and real-world problems. Despite not being previously used for clustering purposes, the impressive performance of COOT in some tests compared to the K-Means algorithm showcases its success and potential to pioneer different studies aimed at expanding its usage in the clustering domain.

Heterogeneity of glycaemic phenotypes in type 1 diabetes

Aims/hypothesisOur study aims to uncover glycaemic phenotype heterogeneity in type 1 diabetes.MethodsIn the Study of the French-speaking Society of Type 1 Diabetes (SFDT1), we characterised glycaemic heterogeneity thanks to a set of complementary metrics: HbA1c, time in range (TIR), time below range (TBR), CV, Gold score and glycaemia risk index (GRI). Applying the Discriminative Dimensionality Reduction with Trees (DDRTree) algorithm, we created a phenotypic tree, i.e. a 2D visual mapping. We also carried out a clustering analysis for comparison.ResultsWe included 618 participants with type 1 diabetes (52.9% men, mean age 40.6 years [SD 14.1]). Our phenotypic tree identified seven glycaemic phenotypes. The 2D phenotypic tree comprised a main branch in the proximal region and glycaemic phenotypes in the distal areas. Dimension 1, the horizontal dimension, was positively associated with GRI (coefficient [95% CI]) (0.54 [0.52, 0.57]), HbA1c (0.39 [0.35, 0.42]), CV (0.24 [0.19, 0.28]) and TBR (0.11 [0.06, 0.15]), and negatively with TIR (−0.52 [−0.54, −0.49]). The vertical dimension was positively associated with TBR (0.41 [0.38, 0.44]), CV (0.40 [0.37, 0.43]), TIR (0.16 [0.12, 0.20]), Gold score (0.10 [0.06, 0.15]) and GRI (0.06 [0.02, 0.11]), and negatively with HbA1c (−0.21 [−0.25, −0.17]). Notably, socioeconomic factors, cardiovascular risk indicators, retinopathy and treatment strategy were significant determinants of glycaemic phenotype diversity. The phenotypic tree enabled more granularity than traditional clustering in revealing clinically relevant subgroups of people with type 1 diabetes.Conclusions/interpretationOur study advances the current understanding of the complex glycaemic profile in people with type 1 diabetes and suggests that strategies based on isolated glycaemic metrics might not capture the complexity of the glycaemic phenotypes in real life. Relying on these phenotypes could improve patient stratification in type 1 diabetes care and personalise disease management.Graphical

An axiomatic framework for three-way clustering

Three-way clustering provides a variety of models with richer structural features than traditional two-way clustering. However, the structural properties of three-way clustering have not been systematically studied. In this paper, we propose an axiomatic framework to study three-way clustering based on the structural properties of three-way clusters. We categorize three-way clustering models into 16 types organized in five levels according to six axioms. We propose three strategies for three-way clustering approaches: the 2to3WC Strategy, the I3WC Strategy, and the E3WC Strategy. These strategies comprehensively cover existing three-way clustering models. We examine each of the existing methods and incorporate almost all of them into these three strategies. The framework not only summarizes the structural properties and methods of existing studies but also provides inspiration for future research. There are six types of three-way clustering models that have not yet been proposed and require further study. Strategies that have not been applied to each type also suggest possible future research directions.

Design of Fast Mining Algorithm for Educational Sport Course Data Based on Cluster Analysis

In this age of big data, education researchers are reconceptualizing and re-evaluating the value of education data. Therefore, we need to use educational data mining methods for data analysis to better guide teaching. The informatization level of colleges and universities is improving year by year, and the entire training data of students from enrollment to graduation is stored. These datasets are collected, stored, and kept by different departments, contain a large amount of regular and relevant information, and truly record the growth footprint of students. Traditional educational decision-making has not yet fully explored and used the valuable information hidden in data resources. Although some scholars have carried out research related to campus data mining at this stage, there are still many problems that have not yet been solved in the application of decision-making in colleges and universities. This paper is based on the idea of data-driven decision-making, combined with the data characteristics of campus big data, and establishes a model solution for student behavior analysis and behavior prediction by applying multiple machine learning algorithms. On the basis of the analysis of students’ academic behavior performance in the context of multi-category educational data, we proposed a cluster analysis framework for processing multi-type campus big data, and described the group characteristics of the clustering results. By introducing the K-prototype algorithm, we effectively solved the multi-category problem where traditional clustering algorithms (such as K-Means, etc.) cannot adapt to the attributes of educational data. The research results show that innovative educational decision-making models and methods are based on the idea of “data-prediction-decision”, which promotes the application research of big data science in the area of education.

Multi-Objective Automatic Clustering Algorithm Based on Evolutionary Multi-Tasking Optimization

Data mining technology is the process of extracting hidden knowledge and potentially useful information from a large number of incomplete, noisy, and random practical application data. The clustering algorithm based on multi-objective evolution has obvious advantages compared with the traditional single-objective method. In order to further improve the performance of evolutionary multi-objective clustering algorithms, this paper proposes a multi-objective automatic clustering model based on evolutionary multi-task optimization. Based on the multi-objective clustering algorithm that automatically determines the value of k, evolutionary multi-task optimization is introduced to deal with multiple clustering tasks simultaneously. A set of non-dominated solutions for clustering results is obtained by concurrently optimizing the overall deviation and connectivity index. Multi-task adjacency coding based on a locus adjacency graph was designed to encode the clustered data. Additionally, an evolutionary operator based on relevance learning was designed to facilitate the evolution of individuals within the population. It also facilitates information transfer between individuals with different tasks, effectively avoiding negative transfer. Finally, the proposed algorithm was applied to both artificial datasets and UCI datasets for testing. It was then compared with traditional clustering algorithms and other multi-objective clustering algorithms. The results verify the advantages of the proposed algorithm in clustering accuracy and algorithm convergence.

Feasibility analysis of using CO2 as a building fire detection indicator

CO2 has rarely been examined as a fire indicator owing to the uncertainty in its background concentration. Nevertheless, for a specific building space with a detector, the background CO2 concentration signal can be determined and documented using artificial intelligence technology. Based on the aforementioned concept, this study explored the feasibility of using CO2 signal features as fire indicators for building fire detection. First, CO2 concentrations under fire and no-fire conditions were obtained to generate study databases. Second, the deep temporal clustering (DTC) model was used to classify CO2 signal features between fire and no-fire scenes, and the clustering accuracy and purity were selected to determine the clustering performance. Third, the k-means, agglomerative clustering, and spectral bi-clustering models were employed to validate the clustering advantages of the DTC model. The results proved the CO2 concentration to be more suitable as a combustion indicator for fire detection compared to the CO2 concentration rate, exhibiting higher clustering accuracy and purity values. Additionally, for time-series clustering, the DTC model outperformed traditional clustering models. The findings of this study represent a fundamental step toward the realization of space-exclusive multi-sensor fire-detection technology for assisting smart building construction in a cost-effective and accurate manner.

Preserving Global Information for Graph Clustering with Masked Autoencoders

Graph clustering aims to divide nodes into different clusters without labels and has attracted great attention due to the success of graph neural networks (GNNs). Traditional GNN-based clustering methods are based on the homophilic assumption, i.e., connected nodes belong to the same clusters. However, this assumption is not always true, as heterophilic graphs are also ubiquitous in the real world, which limits the application of GNNs. Furthermore, these methods overlook global positions, which can result in erroneous clustering. To solve the aforementioned problems, we propose a novel model called Preserving Global Information for Graph Clustering with Masked Autoencoders (GCMA). We first propose a low–high-pass filter to capture meaningful low- and high-frequency information. Then, we propose a graph diffusion method to obtain the global position. Specifically, a parameterized Laplacian matrix is proposed to better control the global direction. To further enhance the learning ability of the autoencoders, we design a model with a masking strategy that enhances the learning ability. Extensive experiments on both homophilic and heterophilic graphs demonstrate GCMA’s advantages over state-of-the-art baselines.

Conditional heavy hitter monitoring and application of heterogeneous graph streams based on sketches

Graph streams offer valuable insights into real network evolution in the era of big data. Heavy hitters and conditional heavy hitters focus on identifying entities with significant network traffic, revealing network data characteristics. In this paper, we present two methods, BLSketch and MHLSketch, for identifying heavy hitters and conditional heavy hitters in heterogeneous graph streams. BLSketch uses the bidirectional mapper, Bimap, for efficient query response times while retaining candidate information. MHLSketch uses min heap to preserve a minimal set of candidates to minimize storage space, excelling in space efficiency. The two methods can be incorporated with any sketch based graph summarization method, showing excellent adaptability. Experimental results on four real-world heterogeneous datasets of varying scales with the largest dataset containing millions of edges demonstrate our methods significantly improve the quality of query answering, outperforming state-of-the-art methods, with precision close to 1 and average relative error close to 0. We further combine our methods with density clustering to introduce the SDC model, which efficiently identifies hot regions and key users in Location-Based Social Networks (LBSN). In experiments, SDC outperforms traditional clustering by reducing up to 93.1% of cold Points of Interest (POIs) and shows strong potential for practical applications.

CAVE: Cerebral artery–vein segmentation in digital subtraction angiography

Cerebral X-ray digital subtraction angiography (DSA) is a widely used imaging technique in patients with neurovascular disease, allowing for vessel and flow visualization with high spatio-temporal resolution. Automatic artery–vein segmentation in DSA plays a fundamental role in vascular analysis with quantitative biomarker extraction, facilitating a wide range of clinical applications. The widely adopted U-Net applied on static DSA frames often struggles with disentangling vessels from subtraction artifacts. Further, it falls short in effectively separating arteries and veins as it disregards the temporal perspectives inherent in DSA. To address these limitations, we propose to simultaneously leverage spatial vasculature and temporal cerebral flow characteristics to segment arteries and veins in DSA. The proposed network, coined CAVE, encodes a 2D+time DSA series using spatial modules, aggregates all the features using temporal modules, and decodes it into 2D segmentation maps. On a large multi-center clinical dataset, CAVE achieves a vessel segmentation Dice of 0.84 (±0.04) and an artery–vein segmentation Dice of 0.79 (±0.06). CAVE surpasses traditional Frangi-based k-means clustering (P < 0.001) and U-Net (P < 0.001) by a significant margin, demonstrating the advantages of harvesting spatio-temporal features. This study represents the first investigation into automatic artery–vein segmentation in DSA using deep learning. The code is publicly available at https://github.com/RuishengSu/CAVE_DSA.

Multitask-Guided Deep Clustering With Boundary Adaptation.

Multitask learning uses external knowledge to improve internal clustering and single-task learning. Existing multitask learning algorithms mostly use shallow-level correlation to aid judgment, and the boundary factors on high-dimensional datasets often lead algorithms to poor performance. The initial parameters of these algorithms cause the border samples to fall into a local optimal solution. In this study, a multitask-guided deep clustering (DC) with boundary adaptation (MTDC-BA) based on a convolutional neural network autoencoder (CNN-AE) is proposed. In the first stage, dubbed multitask pretraining (M-train), we construct an autoencoder (AE) named CNN-AE using the DenseNet-like structure, which performs deep feature extraction and stores captured multitask knowledge into model parameters. In the second phase, the parameters of the M-train are shared for CNN-AE, and clustering results are obtained by deep features, which is termed as single-task fitting (S-fit). To eliminate the boundary effect, we use data augmentation and improved self-paced learning to construct the boundary adaptation. We integrate boundary adaptors into the M-train and S-fit stages appropriately. The interpretability of MTDC-BA is accomplished by data transformation. The model relies on the principle that features become important as the reconfiguration loss decreases. Experiments on a series of typical datasets confirm the performance of the proposed MTDC-BA. Compared with other traditional clustering methods, including single-task DC algorithms and the latest multitask clustering algorithms, our MTDC-BA achieves better clustering performance with higher computational efficiency. Deep features clustering results demonstrate the stability of MTDC-BA by visualization and convergence verification. Through the visualization experiment, we explain and analyze the whole model data input and the middle characteristic layer. Further understanding of the principle of MTDC-BA. Through additional experiments, we know that the proposed MTDC-BA is efficient in the use of multitask knowledge. Finally, we carry out sensitivity experiments on the hyper-parameters to verify their optimal performance.

Global aerosol-type classification using a new hybrid algorithm and Aerosol Robotic Network data

Abstract. The properties of aerosols are highly uncertain owing to the complex changes in their composition in different regions. The radiative properties of different aerosol types differ considerably and are vital for studying aerosol regional and/or global climate effects. Traditional aerosol-type identification algorithms, generally based on cluster or empirical analysis methods, are often inaccurate and time-consuming. In response, our study aimed to develop a new aerosol-type classification model using an innovative hybrid algorithm to improve the precision and efficiency of aerosol-type identification. This novel algorithm incorporates an optical database, constructed using the Mie scattering model, and employs a random forest algorithm to classify different aerosol types based on the optical data from the database. The complex refractive index was used as a baseline to assess the performance of our hybrid algorithm against the traditional Gaussian kernel density clustering method for aerosol-type identification. The hybrid algorithm demonstrated impressive consistency rates of 90 %, 85 %, 84 %, 84 %, and 100 % for dust, mixed-coarse (mixed, course-mode aerosol), mixed-fine (mixed, fine-mode aerosol), urban/industrial, and biomass burning aerosols, respectively. Moreover, it achieved remarkable precision, with evaluation metric indexes for micro-precision, micro-recall, micro-F1-score, and accuracy of 95 %, 89 %, 91 %, and 89 %, respectively. Lastly, a global map of aerosol types was generated using the new hybrid algorithm to characterize aerosol types across the five continents. This study, utilizing a novel approach for the classification of aerosol, will help improve the accuracy of aerosol inversion and determine the sources of aerosol pollution.

Cauchy hyper-graph Laplacian nonnegative matrix factorization for single-cell RNA-sequencing data analysis.

Many important biological facts have been found as single-cell RNA sequencing (scRNA-seq) technology has advanced. With the use of this technology, it is now possible to investigate the connections among individual cells, genes, and illnesses. For the analysis of single-cell data, clustering is frequently used. Nevertheless, biological data usually contain a large amount of noise data, and traditional clustering methods are sensitive to noise. However, acquiring higher-order spatial information from the data alone is insufficient. As a result, getting trustworthy clustering findings is challenging. We propose the Cauchy hyper-graph Laplacian non-negative matrix factorization (CHLNMF) as a unique approach to address these issues. In CHLNMF, we replace the measurement based on Euclidean distance in the conventional non-negative matrix factorization (NMF), which can lessen the influence of noise, with the Cauchy loss function (CLF). The model also incorporates the hyper-graph constraint, which takes into account the high-order link among the samples. The CHLNMF model's best solution is then discovered using a half-quadratic optimization approach. Finally, using seven scRNA-seq datasets, we contrast the CHLNMF technique with the other nine top methods. The validity of our technique was established by analysis of the experimental outcomes.

Integration and Optimization of Educational Teaching Resources in Colleges and Universities Based on Clustering Algorithm

The paper presents Centroid Ranking Optimized Clustering (CROC), a novel approach aimed at enhancing educational teaching and resource optimization through advanced clustering techniques. CROC integrates centroid-based clustering with ranking optimization strategies to segment students or educational resources into meaningful clusters, facilitating personalized learning experiences and targeted interventions. Experimental analysis conducted on real-world educational datasets demonstrates that CROC outperforms traditional clustering algorithms in terms of clustering quality, accuracy, and interpretability. The findings suggest that CROC holds significant promise for informing data-driven decision-making in educational settings, enabling educators to better understand student performance patterns, identify at-risk students, and tailor instructional strategies to meet diverse learning needs. Experimental analysis on real-world educational datasets reveals that CROC achieves a Silhouette Score of 0.75 and a Davies-Bouldin Index of 0.40 on the MathScores dataset, outperforming traditional algorithms such as k-means (Silhouette Score: 0.68, Davies-Bouldin Index: 0.53), Hierarchical (Silhouette Score: 0.62, Davies-Bouldin Index: 0.57), and DBSCAN (Silhouette Score: 0.45, Davies-Bouldin Index: 0.78). Similarly, on the EnglishTest dataset, CROC attains a Silhouette Score of 0.82 and a Davies-Bouldin Index of 0.35, surpassing k-means (Silhouette Score: 0.74, Davies-Bouldin Index: 0.47), Hierarchical (Silhouette Score: 0.68, Davies-Bouldin Index: 0.52), and DBSCAN (Silhouette Score: 0.53, Davies-Bouldin Index: 0.68). These findings underscore the effectiveness of CROC in clustering educational data, enabling personalized learning experiences and data-driven decision-making in educational settings.

Multivariate time series clustering based on fuzzy cognitive maps and community detection

Most time series clustering methods mainly focus on univariate time series (UTS). Compared with UTS, multivariate time series (MTS) consists of multiple components. Although interest in MTS clustering is increasing, its performance is far from satisfactory. Most traditional MTS clustering methods may have two limitations. First, they do not consider both the temporal features of each component and the relationship between the components. Second, they can only identify the local relationship between adjacent data, but cannot obtain long-distance global relationships and capture clusters of arbitrary shapes. In this paper, we develop a method for MTS clustering based on fuzzy cognitive maps (FCMs) and community detection, termed as MTSC-FCM-CD. To overcome the first limitation, we use FCM to represent MTS; FCM can extract temporal features while preserving the relationship between components in MTS. To overcome the second limitation, we use the community detection algorithm to cluster the global relations, which is different from the traditional nearest neighbor distance-based method. In the calculation process, the similarity between FCMs should be used to build a complex network. Existing methods calculate the similarity between FCMs, only considering the numerical characteristics but ignoring the topological structure. To solve this problem, we design a two-stage similarity measure to build a complex network. In comparison to the existing methods, the experimental results on twenty benchmark datasets demonstrate the effectiveness of MTSC-FCM-CD in MTS clustering.

One-step incremental multi-view spectral clustering based on graph linkage learning

Most traditional multi-view spectral clustering algorithms involve two separate steps: solving the spectral embedding matrix and clustering, which may introduce errors during the clustering process. Moreover, in practical applications, the number of available views may increase over time, and the approach of re-fusing all views in each computation would result in elevated computational costs. In this paper, we propose a novel model: One-step Incremental Multi-view Spectral Clustering based on Graph Linkage learning (OIMvSC). OIMvSC only needs to store the consensus spectral embedding matrix and the consensus label matrix of all previous views and combines them with the spectral embedding matrix of the newly available view to solve the fused consensus spectral embedding matrix and consensus label matrix. To further enhance clustering performance, OIMvSC introduces graph linkage learning, which reduces erroneous connections between clusters while preserving correct connections within clusters. A convergent iterative algorithm for solving OIMvSC is proposed. Experimental results demonstrate that OIMvSC exhibits excellent clustering performance.

A secure routing protocol using trust‐based clustering and bionic intelligence algorithm for UAV‐assisted vehicular ad hoc networks

AbstractVehicular ad hoc networks (VANET) are one of the advanced technologies for distributing dynamic vehicular information across the globe. The VANET is extensively used in many applications, especially road safety applications and intelligent transport systems (ITS). However, direct communication causes high bandwidth (BW) requirement and power consumption. Hence, this article introduces a clustering‐based mechanism to communicate vehicles with infrastructures. The cluster head (CH) is formed based on certain rules, and nodes or vehicles are combined. But, maintaining stability remains challenging for the traditional clustering mechanism. Moreover, the developed technique must examine the malicious and reduce the risk of fake information sharing. This research emphasizes a trust‐based clustering mechanism to select the CH based on a vehicle's knowledge, reputation, and experience. In addition to this, the backup head is also analyzed to promote trust in each vehicle. After clustering, secure routing is undertaken. For this, a bionic remora optimization algorithm (BROA) is proposed, and it considers the hop Count Field as well as the transmission range of vehicles to select the best routes. The performance measures such as end‐to‐end delay ratio, packet delivery ratio (PDR), throughput, trust values, energy consumption are analyzed and compared with existing techniques. In an experimental scenario, the proposed technique has an end‐to‐end delay of 3.8 ms, PDR of 98%, trust value of .4 and .06 for wormhole and Sybil attack, energy of and throughput of 78.7 kbps are attained. The outcome results prove the efficacy of a proposed method.

Survey on graph neural network-based community detection and its applications

Detecting communities within a network is a critical component of network analysis. The process involves identifying clusters of nodes that exhibit greater similarity to each other compared to other nodes in the network. In the context of Complex networks (CN), community detection becomes even more important as these clusters provide relevant information of interest. Traditional mathematical and clustering methods have limitations in terms of data visualization and high-dimensional information extraction. To address these challenges, graph neural network learning methods have gained popularity in community detection, as they are capable of handling complex structures and multi-dimensional data. Developing a framework for community detection in complex networks using graph neural network learning is a challenging and ongoing research objective. Therefore, it is essential for researchers to conduct a thorough review of community detection techniques that utilize cutting-edge graph neural network learning methods [102], in order to analyze and construct effective detection models. This paper provides a brief overview of graph neural network learning methods based on community detection methods and summarizes datasets, evaluation metrics, applications, and challenges of community detection in complex networks.

Near-neighbor Propagation Clustering Algorithm Based on Cuckoo Search

In this paper, a nearest neighbor propagation clustering algorithm (CSB-AP) based on cuckoo search is proposed to solve the problem of poor parameter setting of the AP algorithm. A population-intelligent optimized cuckoo search algorithm is introduced into the AP algorithm to find the appropriate parameters. Two important parameters in the nearest neighbor propagation algorithm are taken as the location of the bird’s nest. The BWP value index is introduced as the bird’s nest fitness in the process of intelligent search, and the minimum value is obtained as the final fitness through the reverse mechanism. The final parameters are substituted into the calculation as the best parameters. The CSB-AP algorithm is verified by the UCI data set, and compared with the traditional AP nearest neighbor propagation clustering algorithm, it is found that the CSB-AP algorithm proposed in this paper is better than the traditional AP nearest neighbor propagation clustering algorithm. By comparison, it can be found that the clustering result obtained by the CSB-AP algorithm is closer to the actual result, and it can be known according to the results of the BWP index, contour coefficient, Recall, and F-measure. The improved algorithm can significantly improve the clustering quality and clustering performance.