In recent years,there has been a concerted effort to improve anomaly detection tech-niques,particularly in the context of high-dimensional,distributed clinical data.Analysing patient data within clinical settings reve...In recent years,there has been a concerted effort to improve anomaly detection tech-niques,particularly in the context of high-dimensional,distributed clinical data.Analysing patient data within clinical settings reveals a pronounced focus on refining diagnostic accuracy,personalising treatment plans,and optimising resource allocation to enhance clinical outcomes.Nonetheless,this domain faces unique challenges,such as irregular data collection,inconsistent data quality,and patient-specific structural variations.This paper proposed a novel hybrid approach that integrates heuristic and stochastic methods for anomaly detection in patient clinical data to address these challenges.The strategy combines HPO-based optimal Density-Based Spatial Clustering of Applications with Noise for clustering patient exercise data,facilitating efficient anomaly identification.Subsequently,a stochastic method based on the Interquartile Range filters unreliable data points,ensuring that medical tools and professionals receive only the most pertinent and accurate information.The primary objective of this study is to equip healthcare pro-fessionals and researchers with a robust tool for managing extensive,high-dimensional clinical datasets,enabling effective isolation and removal of aberrant data points.Furthermore,a sophisticated regression model has been developed using Automated Machine Learning(AutoML)to assess the impact of the ensemble abnormal pattern detection approach.Various statistical error estimation techniques validate the efficacy of the hybrid approach alongside AutoML.Experimental results show that implementing this innovative hybrid model on patient rehabilitation data leads to a notable enhance-ment in AutoML performance,with an average improvement of 0.041 in the R2 score,surpassing the effectiveness of traditional regression models.展开更多
Cluster analysis is a crucial technique in unsupervised machine learning,pattern recognition,and data analysis.However,current clustering algorithms suffer from the need for manual determination of parameter values,lo...Cluster analysis is a crucial technique in unsupervised machine learning,pattern recognition,and data analysis.However,current clustering algorithms suffer from the need for manual determination of parameter values,low accuracy,and inconsistent performance concerning data size and structure.To address these challenges,a novel clustering algorithm called the fully automated density-based clustering method(FADBC)is proposed.The FADBC method consists of two stages:parameter selection and cluster extraction.In the first stage,a proposed method extracts optimal parameters for the dataset,including the epsilon size and a minimum number of points thresholds.These parameters are then used in a density-based technique to scan each point in the dataset and evaluate neighborhood densities to find clusters.The proposed method was evaluated on different benchmark datasets andmetrics,and the experimental results demonstrate its competitive performance without requiring manual inputs.The results show that the FADBC method outperforms well-known clustering methods such as the agglomerative hierarchical method,k-means,spectral clustering,DBSCAN,FCDCSD,Gaussian mixtures,and density-based spatial clustering methods.It can handle any kind of data set well and perform excellently.展开更多
Clustering is used to gain an intuition of the struc tures in the data.Most of the current clustering algorithms pro duce a clustering structure even on data that do not possess such structure.In these cases,the algor...Clustering is used to gain an intuition of the struc tures in the data.Most of the current clustering algorithms pro duce a clustering structure even on data that do not possess such structure.In these cases,the algorithms force a structure in the data instead of discovering one.To avoid false structures in the relations of data,a novel clusterability assessment method called density-based clusterability measure is proposed in this paper.I measures the prominence of clustering structure in the data to evaluate whether a cluster analysis could produce a meaningfu insight to the relationships in the data.This is especially useful in time-series data since visualizing the structure in time-series data is hard.The performance of the clusterability measure is evalu ated against several synthetic data sets and time-series data sets which illustrate that the density-based clusterability measure can successfully indicate clustering structure of time-series data.展开更多
Clustering evolving data streams is important to be performed in a limited time with a reasonable quality. The existing micro clustering based methods do not consider the distribution of data points inside the micro c...Clustering evolving data streams is important to be performed in a limited time with a reasonable quality. The existing micro clustering based methods do not consider the distribution of data points inside the micro cluster. We propose LeaDen-Stream (Leader Density-based clustering algorithm over evolving data Stream), a density-based clustering algorithm using leader clustering. The algorithm is based on a two-phase clustering. The online phase selects the proper mini-micro or micro-cluster leaders based on the distribution of data points in the micro clusters. Then, the leader centers are sent to the offline phase to form final clusters. In LeaDen-Stream, by carefully choosing between two kinds of micro leaders, we decrease time complexity of the clustering while maintaining the cluster quality. A pruning strategy is also used to filter out real data from noise by introducing dense and sparse mini-micro and micro-cluster leaders. Our performance study over a number of real and synthetic data sets demonstrates the effectiveness and efficiency of our method.展开更多
Encephalitis is a brain inflammation disease.Encephalitis can yield to seizures,motor disability,or some loss of vision or hearing.Sometimes,encepha-litis can be a life-threatening and proper diagnosis in an early stag...Encephalitis is a brain inflammation disease.Encephalitis can yield to seizures,motor disability,or some loss of vision or hearing.Sometimes,encepha-litis can be a life-threatening and proper diagnosis in an early stage is very crucial.Therefore,in this paper,we are proposing a deep learning model for computerized detection of Encephalitis from the electroencephalogram data(EEG).Also,we propose a Density-Based Clustering model to classify the distinctive waves of Encephalitis.Customary clustering models usually employ a computed single centroid virtual point to define the cluster configuration,but this single point does not contain adequate information.To precisely extract accurate inner structural data,a multiple centroids approach is employed and defined in this paper,which defines the cluster configuration by allocating weights to each state in the cluster.The multiple EEG view fuzzy learning approach incorporates data from every sin-gle view to enhance the model's clustering performance.Also a fuzzy Density-Based Clustering model with multiple centroids(FDBC)is presented.This model employs multiple real state centroids to define clusters using Partitioning Around Centroids algorithm.The Experimental results validate the medical importance of the proposed clustering model.展开更多
With the rapid advance of wireless communication, tracking the positions of the moving objects is becoming increasingly feasible and necessary. Because a large number of people use mobile phones, we must handle a larg...With the rapid advance of wireless communication, tracking the positions of the moving objects is becoming increasingly feasible and necessary. Because a large number of people use mobile phones, we must handle a large moving object database as well as the following problems. How can we provide the customers with high quality service, that means, how can we deal with so many enquiries within as less time as possible? Because of the large number of data, the gap between CPU speed and the size of main memory has increasing considerably. One way to reduce the time to handle enquiries is to reduce the I/O number between the buffer and the secondary storage.An effective clustering of the objects can minimize the I/O cost between them. In this paper, according to the characteristic of the moving object database, we analyze the objects in buffer, according to their mappings in the two dimension coordinate, and then develop a density based clustering method to effectively reorganize the clusters. This new mechanism leads to the less cost of the I/O operation and the more efficient response to enquiries.展开更多
Finding clusters based on density represents a significant class of clustering algorithms.These methods can discover clusters of various shapes and sizes.The most studied algorithm in this class is theDensity-Based Sp...Finding clusters based on density represents a significant class of clustering algorithms.These methods can discover clusters of various shapes and sizes.The most studied algorithm in this class is theDensity-Based Spatial Clustering of Applications with Noise(DBSCAN).It identifies clusters by grouping the densely connected objects into one group and discarding the noise objects.It requires two input parameters:epsilon(fixed neighborhood radius)and MinPts(the lowest number of objects in epsilon).However,it can’t handle clusters of various densities since it uses a global value for epsilon.This article proposes an adaptation of the DBSCAN method so it can discover clusters of varied densities besides reducing the required number of input parameters to only one.Only user input in the proposed method is the MinPts.Epsilon on the other hand,is computed automatically based on statistical information of the dataset.The proposed method finds the core distance for each object in the dataset,takes the average of these distances as the first value of epsilon,and finds the clusters satisfying this density level.The remaining unclustered objects will be clustered using a new value of epsilon that equals the average core distances of unclustered objects.This process continues until all objects have been clustered or the remaining unclustered objects are less than 0.006 of the dataset’s size.The proposed method requires MinPts only as an input parameter because epsilon is computed from data.Benchmark datasets were used to evaluate the effectiveness of the proposed method that produced promising results.Practical experiments demonstrate that the outstanding ability of the proposed method to detect clusters of different densities even if there is no separation between them.The accuracy of the method ranges from 92%to 100%for the experimented datasets.展开更多
We propose a new clustering algorithm that assists the researchers to quickly and accurately analyze data. We call this algorithm Combined Density-based and Constraint-based Algorithm (CDC). CDC consists of two phases...We propose a new clustering algorithm that assists the researchers to quickly and accurately analyze data. We call this algorithm Combined Density-based and Constraint-based Algorithm (CDC). CDC consists of two phases. In the first phase, CDC employs the idea of density-based clustering algorithm to split the original data into a number of fragmented clusters. At the same time, CDC cuts off the noises and outliers. In the second phase, CDC employs the concept of K-means clustering algorithm to select a greater cluster to be the center. Then, the greater cluster merges some smaller clusters which satisfy some constraint rules. Due to the merged clusters around the center cluster, the clustering results show high accuracy. Moreover, CDC reduces the calculations and speeds up the clustering process. In this paper, the accuracy of CDC is evaluated and compared with those of K-means, hierarchical clustering, and the genetic clustering algorithm (GCA) proposed in 2004. Experimental results show that CDC has better performance.展开更多
Density-based clustering is an important category among clustering algorithms. In real applications, many datasets suffer from incompleteness. Traditional imputation technologies or other techniques for handling missi...Density-based clustering is an important category among clustering algorithms. In real applications, many datasets suffer from incompleteness. Traditional imputation technologies or other techniques for handling missing values are not suitable for density-based clustering and decrease clustering result quality. To avoid these problems,we develop a novel density-based clustering approach for incomplete data based on Bayesian theory, which conducts imputation and clustering concurrently and makes use of intermediate clustering results. To avoid the impact of low-density areas inside non-convex clusters, we introduce a local imputation clustering algorithm, which aims to impute points to high-density local areas. The performances of the proposed algorithms are evaluated using ten synthetic datasets and five real-world datasets with induced missing values. The experimental results show the effectiveness of the proposed algorithms.展开更多
Ball milling is widely used in industry to mill particulate material.The primary purpose of this process is to attain an appropriate product size with the least possible energy consumption.The process is also extensiv...Ball milling is widely used in industry to mill particulate material.The primary purpose of this process is to attain an appropriate product size with the least possible energy consumption.The process is also extensively utilised in pharmaceuticals for the comminution of the excipients or drugs.Surprisingly,for ball mill,little is known concerning the mechanism of size reduction.Traditional prediction approaches are not deemed useful to provide significant insights into the operation or facilitate radical step changes in performance.Therefore,the discrete element method(DEM)as a computational modelling approach has been used in this paper.In previous research,DEM has been applied to simulate breaking behaviour through the impact energy of all ball collisions as the driving force for fracturing.However,the nature of pharmaceutical material fragmentation during ball milling is more complex.Suitable functional equations which link broken media and applied energy do not consider the collision of particulate media of different shapes or collisions of particulate media(such as granules)with balls and rotating mill drum.This could have a significant impact on fragmentation.Therefore,this paper aimed to investigate the fragmentation of bounded particles into DEM granules of different shape/size during the ball milling process.A systematic study was undertaken to explore the effect of milling speed on breakage behaviour.Also,in this study,a combination of a density-based clustering method and discrete element method was employed to numerically investigate the number and size of the fragments generated during the ball milling process over time.It was discovered that the collisions of the ball increased proportionally with rotation speed until reaching the critical rotation speed.Consequently,results illustrate that with an increase of rotation speed,the mill power increased correspondingly.The caratacting motion of mill material together with balls was identified as the most effective regime regarding the fragmentation,and fewer breakage events occurred for centrifugal motion.Higher quantities of the fines in each batch were produced with increased milling speed with less quantities of grain fragments.Moreover,the relationship between the number of produced fragment and milling speed at the end of the process exhibited a linear tendency.展开更多
Overlapping community detection in a network is a challenging issue which attracts lots of attention in recent years.A notion of hesitant node（HN） is proposed. An HN contacts with multiple communities while the comm...Overlapping community detection in a network is a challenging issue which attracts lots of attention in recent years.A notion of hesitant node（HN） is proposed. An HN contacts with multiple communities while the communications are not strong or even accidental, thus the HN holds an implicit community structure.However, HNs are not rare in the real world network. It is important to identify them because they can be efficient hubs which form the overlapping portions of communities or simple attached nodes to some communities. Current approaches have difficulties in identifying and clustering HNs. A density-based rough set model（DBRSM） is proposed by combining the virtue of densitybased algorithms and rough set models. It incorporates the macro perspective of the community structure of the whole network and the micro perspective of the local information held by HNs, which would facilitate the further ＂growth＂ of HNs in community. We offer a theoretical support for this model from the point of strength of the trust path. The experiments on the real-world and synthetic datasets show the practical significance of analyzing and clustering the HNs based on DBRSM. Besides, the clustering based on DBRSM promotes the modularity optimization.展开更多
Under hydrothermal and solvothermal conditions,two novel cobalt-based complexes,{[Co_(2)(CIA)(OH)(1,4-dtb)]·3.2H_(2)O}n(HU23)and{[Co_(2)(CIA)(OH)(1,4-dib)]·3.5H2O·DMF}n(HU24),were successfully construct...Under hydrothermal and solvothermal conditions,two novel cobalt-based complexes,{[Co_(2)(CIA)(OH)(1,4-dtb)]·3.2H_(2)O}n(HU23)and{[Co_(2)(CIA)(OH)(1,4-dib)]·3.5H2O·DMF}n(HU24),were successfully constructed by coordinatively assembling the semi-rigid multidentate ligand 5-(1-carboxyethoxy)isophthalic acid(H₃CIA)with the Nheterocyclic ligands 1,4-di(4H-1,2,4-triazol-4-yl)benzene(1,4-dtb)and 1,4-di(1H-imidazol-1-yl)benzene(1,4-dib),respectively,around Co^(2+)ions.Single-crystal X-ray diffraction analysis revealed that in both complexes HU23 and HU24,the CIA^(3-)anions adopt aκ^(7)-coordination mode,bridging six Co^(2+)ions via their five carboxylate oxygen atoms and one ether oxygen atom.This linkage forms tetranuclear[Co4(μ3-OH)2]^(6+)units.These Co-oxo cluster units were interconnected by CIA^(3-)anions to assemble into 2D kgd-type structures featuring a 3,6-connected topology.The 2D layers were further connected by 1,4-dtb and 1,4-dib,resulting in 3D pillar-layered frameworks for HU23 and HU24.Notably,despite the similar configurations of 1,4-dtb and 1,4-dib,differences in their coordination spatial orientations lead to topological divergence in the 3D frameworks of HU23 and HU24.Topological analysis indicates that the frameworks of HU23 and HU24 can be simplified into a 3,10-connected net(point symbol:(4^(10).6^(3).8^(2))(4^(3))_(2))and a 3,8-connected tfz-d net(point symbol:(4^(3))_(2)((4^(6).6^(18).8^(4)))),respectively.This structural differentiation confirms the precise regulatory role of ligands on the topology of metal-organic frameworks.Moreover,the ultraviolet-visible absorption spectra confirmed that HU23 and HU24 have strong absorption capabilities for ultraviolet and visible light.According to the Kubelka-Munk method,their bandwidths were 2.15 and 2.08 eV,respectively,which are consistent with those of typical semiconductor materials.Variable-temperature magnetic susceptibility measurements(2-300 K)revealed significant antiferromagnetic coupling in both complexes,with their effective magnetic moments decreasing markedly as the temperature lowered.CCDC:2457554,HU23;2457553,HU24.展开更多
As large-scale astronomical surveys,such as the Sloan Digital Sky Survey(SDSS)and the Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST),generate increasingly complex datasets,clustering algorithms have...As large-scale astronomical surveys,such as the Sloan Digital Sky Survey(SDSS)and the Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST),generate increasingly complex datasets,clustering algorithms have become vital for identifying patterns and classifying celestial objects.This paper systematically investigates the application of five main categories of clustering techniques-partition-based,density-based,model-based,hierarchical,and“others”-across a range of astronomical research over the past decade.This review focuses on the six key application areas of stellar classification,galaxy structure analysis,detection of galactic and interstellar features,highenergy astrophysics,exoplanet studies,and anomaly detection.This paper provides an in-depth analysis of the performance and results of each method,considering their respective suitabilities for different data types.Additionally,it presents clustering algorithm selection strategies based on the characteristics of the spectroscopic data being analyzed.We highlight challenges such as handling large datasets,the need for more efficient computational tools,and the lack of labeled data.We also underscore the potential of unsupervised and semi-supervised clustering approaches to overcome these challenges,offering insight into their practical applications,performance,and results in astronomical research.展开更多
Bottom-up and top-down endogenous automobile clusters exhibit distinct evolutionary traits and driving mechanisms,yet their comparative analysis remains understudied.Therefore,using Taizhou automobile industry cluster...Bottom-up and top-down endogenous automobile clusters exhibit distinct evolutionary traits and driving mechanisms,yet their comparative analysis remains understudied.Therefore,using Taizhou automobile industry cluster(TAIC)and Wuhu automobile industry cluster(WAIC)as cases,using historical statistical data and field interview data from the 1980s to 2023,combined with qualitative research methods of thematic and diachronic analysis,and quantitative research methods of social network analysis,we compare both endogenous automobile clusters’evolutionary traits and driving mechanisms.The results confirm both clusters undergo multi-scale spatial reconfiguration,organizational complexification,and intelligent networking technological transformation,yet diverge fundamentally:TAIC evolves through market-driven progressive expansion,transitioning from single to dual-core structures via private enterprise networking,with innovation following market-integrated logic and institutional thickness built on demand-driven evolution.Conversely,WAIC follows planned expansion,maintaining state-led hierarchical single-core stability through policy-driven breakthrough innovation and supply-dominated institutional construction-though both ultimately require formal-informal system synergy.Their coevolution is driven by dynamic interactions of path dependence(weakening influence),learning-innovation(strengthening influence),and relationship selection(inverted U-shaped trajectory),with divergent development paths rooted in TAIC’s grassroots self-organization genes versus WAIC’s top-level design genes,amplified by core enterprises’strategic disparities.The research findings can not only provide decision-making support for China’s industrial upgrading,but also contribute China’s insights to global economic governance.展开更多
This paper proposes an equivalent modeling method for photovoltaic(PV)power stations via a particle swarm optimization(PSO)K-means clustering(KMC)algorithm with passive filter parameter clustering to address the compl...This paper proposes an equivalent modeling method for photovoltaic(PV)power stations via a particle swarm optimization(PSO)K-means clustering(KMC)algorithm with passive filter parameter clustering to address the complexities,simulation time cost and convergence problems of detailed PV power station models.First,the amplitude–frequency curves of different filter parameters are analyzed.Based on the results,a grouping parameter set for characterizing the external filter characteristics is established.These parameters are further defined as clustering parameters.A single PV inverter model is then established as a prerequisite foundation.The proposed equivalent method combines the global search capability of PSO with the rapid convergence of KMC,effectively overcoming the tendency of KMC to become trapped in local optima.This approach enhances both clustering accuracy and numerical stability when determining equivalence for PV inverter units.Using the proposed clustering method,both a detailed PV power station model and an equivalent model are developed and compared.Simulation and hardwarein-loop(HIL)results based on the equivalent model verify that the equivalent method accurately represents the dynamic characteristics of PVpower stations and adapts well to different operating conditions.The proposed equivalent modeling method provides an effective analysis tool for future renewable energy integration research.展开更多
Villager Pan Chunlin is witnessing a boom in his homestay business.More and more visitors are coming to his village,Yucun Village in Anji County,Huzhou City,Zhejiang Province.
The shift toward specialized and large-scale agricultural production has spurred the emergence of agricultural clusters as key forces of rural vitalization and sustainable development.This paper explored the formation...The shift toward specialized and large-scale agricultural production has spurred the emergence of agricultural clusters as key forces of rural vitalization and sustainable development.This paper explored the formation and evolution of Meizhou pomelo industry cluster in China,focusing on its role in restructuring rural socio-economic systems and integrating the whole value chains.Based on a case study employing qualitative methods such as in-depth interviews and participatory observation,the agricultural cluster evolution of Meizhou pomelo was categorized into three key phases of initial decentralization,self-organized scaling,and reorganized clustering.Geographical proximity and industrial agglomeration constitute the physical foundation,while vertical/horizontal linkages,technologic-al innovation,and policy support enhance competitiveness.Special mechanisms emerge through localized social networks,farmer co-operatives’activation,and cross-regional market expansion.The cluster’s impact is manifested in the shift from extensive to standard-ized and modernized production,diversified and flexible livelihood of farmers,and the integration of agriculture with industry and ser-vices.The development of the whole value chain based on agricultural cluster represents a critical pathway for achieving agricultural modernization,encompassing both internal and external value chain optimization.Through quality assurance systems,product diversi-fication strategies,operational efficiency improvements,and brand enhancement,these clusters amplify product value propositions and market competitiveness.This systemic approach facilitates supply-demand coordination,enables resource synergies,and optimizes eco-nomic returns across the horizontal and vertical value chain.This paper argues that agricultural clusters serve as strategic catalysts for sustainable rural development by reconstructing local production systems,fostering innovation ecosystems,and aligning agricultural modernization.It contributes to debates on rural vitalization by demonstrating how agricultural clustering can reconfigure rural areas as hubs of ecological modernization,rather than mere urban peripheries.展开更多
By encouraging collaboration,attracting young entrepreneurs,and setting up village-invested enterprises,rural communities in Zhejiang Province are growing their income streams.“Over the past years,we have seen more a...By encouraging collaboration,attracting young entrepreneurs,and setting up village-invested enterprises,rural communities in Zhejiang Province are growing their income streams.“Over the past years,we have seen more and more tourists coming to our village,and their stay here has grown longer.Many even said they don’t want to leave,”Pan Chunlin,a resident of Yucun Village in Anji County of Huzhou City,east China’s Zhejiang Province,told China Today.Pan is the owner of the Chunlin Lodge,a bed&breakfast(B&B)which received more than 70,000 visitors last year,generating over RMB 4 million in revenue.展开更多
Noncohesive particle clusters are identified and tracked in turbulent flows to determine the breakdown and time evolution of cluster statistics and their implications for interscale mass transfer,which has connections...Noncohesive particle clusters are identified and tracked in turbulent flows to determine the breakdown and time evolution of cluster statistics and their implications for interscale mass transfer,which has connections to the classical turbulent energy cascade and its mass cascade counterpart running in parallel.In particular,the formation and dynamics of sediment and larvae clusters are of interest to coral larvae settlement in coastal regions and particularly the resilience of green-gray coastal protection solutions.Analogous cluster behavior is relevant to cloud microphysics and precipitation initiation,radiation transport and light transmission through colloids and suspensions,heat and mass transfer in particle-laden flows,and viral and pollutant transmission.Following a comparison between various clustering techniques,we adopt a density-based cluster identification algorithm based on its simplicity and efficiency,where particles are clustered based on the number of neighboring particles in their individual spheres of influence.We establish parallels with lattice-based percolation theory,as evident in the power-law scaling of the cluster size distribution near the percolation threshold.The degree of discontinuity of the phase transition associated with this percolation threshold is observed to broaden with larger Stokes numbers and thereby large-scale clustering.The sensitivity of our findings to the employed clustering algorithm is discussed.A novel cluster tracking algorithm is deployed to determine the interscale transfer rate along the particle-number phase-space dimension via accounting of cluster breakup and merger events,extending previous work on the bubble breakup cascade beneath surface breaking waves.Our findings shed light on the interaction between particle clusters and their carrier turbulent flows,with an eye toward transport models incorporating cluster characteristics and dynamics.展开更多
AIM:To evaluate long-term visual field(VF)prediction using K-means clustering in patients with primary open angle glaucoma(POAG).METHODS:Patients who underwent 24-2 VF tests≥10 were included in this study.Using 52 to...AIM:To evaluate long-term visual field(VF)prediction using K-means clustering in patients with primary open angle glaucoma(POAG).METHODS:Patients who underwent 24-2 VF tests≥10 were included in this study.Using 52 total deviation values(TDVs)from the first 10 VF tests of the training dataset,VF points were clustered into several regions using the hierarchical ordered partitioning and collapsing hybrid(HOPACH)and K-means clustering.Based on the clustering results,a linear regression analysis was applied to each clustered region of the testing dataset to predict the TDVs of the 10th VF test.Three to nine VF tests were used to predict the 10th VF test,and the prediction errors(root mean square error,RMSE)of each clustering method and pointwise linear regression(PLR)were compared.RESULTS:The training group consisted of 228 patients(mean age,54.20±14.38y;123 males and 105 females),and the testing group included 81 patients(mean age,54.88±15.22y;43 males and 38 females).All subjects were diagnosed with POAG.Fifty-two VF points were clustered into 11 and nine regions using HOPACH and K-means clustering,respectively.K-means clustering had a lower prediction error than PLR when n=1:3 and 1:4(both P≤0.003).The prediction errors of K-means clustering were lower than those of HOPACH in all sections(n=1:4 to 1:9;all P≤0.011),except for n=1:3(P=0.680).PLR outperformed K-means clustering only when n=1:8 and 1:9(both P≤0.020).CONCLUSION:K-means clustering can predict longterm VF test results more accurately in patients with POAG with limited VF data.展开更多
文摘In recent years,there has been a concerted effort to improve anomaly detection tech-niques,particularly in the context of high-dimensional,distributed clinical data.Analysing patient data within clinical settings reveals a pronounced focus on refining diagnostic accuracy,personalising treatment plans,and optimising resource allocation to enhance clinical outcomes.Nonetheless,this domain faces unique challenges,such as irregular data collection,inconsistent data quality,and patient-specific structural variations.This paper proposed a novel hybrid approach that integrates heuristic and stochastic methods for anomaly detection in patient clinical data to address these challenges.The strategy combines HPO-based optimal Density-Based Spatial Clustering of Applications with Noise for clustering patient exercise data,facilitating efficient anomaly identification.Subsequently,a stochastic method based on the Interquartile Range filters unreliable data points,ensuring that medical tools and professionals receive only the most pertinent and accurate information.The primary objective of this study is to equip healthcare pro-fessionals and researchers with a robust tool for managing extensive,high-dimensional clinical datasets,enabling effective isolation and removal of aberrant data points.Furthermore,a sophisticated regression model has been developed using Automated Machine Learning(AutoML)to assess the impact of the ensemble abnormal pattern detection approach.Various statistical error estimation techniques validate the efficacy of the hybrid approach alongside AutoML.Experimental results show that implementing this innovative hybrid model on patient rehabilitation data leads to a notable enhance-ment in AutoML performance,with an average improvement of 0.041 in the R2 score,surpassing the effectiveness of traditional regression models.
基金the Deanship of Scientific Research at Umm Al-Qura University,Grant Code:(23UQU4361009DSR001).
文摘Cluster analysis is a crucial technique in unsupervised machine learning,pattern recognition,and data analysis.However,current clustering algorithms suffer from the need for manual determination of parameter values,low accuracy,and inconsistent performance concerning data size and structure.To address these challenges,a novel clustering algorithm called the fully automated density-based clustering method(FADBC)is proposed.The FADBC method consists of two stages:parameter selection and cluster extraction.In the first stage,a proposed method extracts optimal parameters for the dataset,including the epsilon size and a minimum number of points thresholds.These parameters are then used in a density-based technique to scan each point in the dataset and evaluate neighborhood densities to find clusters.The proposed method was evaluated on different benchmark datasets andmetrics,and the experimental results demonstrate its competitive performance without requiring manual inputs.The results show that the FADBC method outperforms well-known clustering methods such as the agglomerative hierarchical method,k-means,spectral clustering,DBSCAN,FCDCSD,Gaussian mixtures,and density-based spatial clustering methods.It can handle any kind of data set well and perform excellently.
文摘Clustering is used to gain an intuition of the struc tures in the data.Most of the current clustering algorithms pro duce a clustering structure even on data that do not possess such structure.In these cases,the algorithms force a structure in the data instead of discovering one.To avoid false structures in the relations of data,a novel clusterability assessment method called density-based clusterability measure is proposed in this paper.I measures the prominence of clustering structure in the data to evaluate whether a cluster analysis could produce a meaningfu insight to the relationships in the data.This is especially useful in time-series data since visualizing the structure in time-series data is hard.The performance of the clusterability measure is evalu ated against several synthetic data sets and time-series data sets which illustrate that the density-based clusterability measure can successfully indicate clustering structure of time-series data.
文摘Clustering evolving data streams is important to be performed in a limited time with a reasonable quality. The existing micro clustering based methods do not consider the distribution of data points inside the micro cluster. We propose LeaDen-Stream (Leader Density-based clustering algorithm over evolving data Stream), a density-based clustering algorithm using leader clustering. The algorithm is based on a two-phase clustering. The online phase selects the proper mini-micro or micro-cluster leaders based on the distribution of data points in the micro clusters. Then, the leader centers are sent to the offline phase to form final clusters. In LeaDen-Stream, by carefully choosing between two kinds of micro leaders, we decrease time complexity of the clustering while maintaining the cluster quality. A pruning strategy is also used to filter out real data from noise by introducing dense and sparse mini-micro and micro-cluster leaders. Our performance study over a number of real and synthetic data sets demonstrates the effectiveness and efficiency of our method.
基金funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2022R113)Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘Encephalitis is a brain inflammation disease.Encephalitis can yield to seizures,motor disability,or some loss of vision or hearing.Sometimes,encepha-litis can be a life-threatening and proper diagnosis in an early stage is very crucial.Therefore,in this paper,we are proposing a deep learning model for computerized detection of Encephalitis from the electroencephalogram data(EEG).Also,we propose a Density-Based Clustering model to classify the distinctive waves of Encephalitis.Customary clustering models usually employ a computed single centroid virtual point to define the cluster configuration,but this single point does not contain adequate information.To precisely extract accurate inner structural data,a multiple centroids approach is employed and defined in this paper,which defines the cluster configuration by allocating weights to each state in the cluster.The multiple EEG view fuzzy learning approach incorporates data from every sin-gle view to enhance the model's clustering performance.Also a fuzzy Density-Based Clustering model with multiple centroids(FDBC)is presented.This model employs multiple real state centroids to define clusters using Partitioning Around Centroids algorithm.The Experimental results validate the medical importance of the proposed clustering model.
基金This work is supported by University IT Research Center Project in KOREA.
文摘With the rapid advance of wireless communication, tracking the positions of the moving objects is becoming increasingly feasible and necessary. Because a large number of people use mobile phones, we must handle a large moving object database as well as the following problems. How can we provide the customers with high quality service, that means, how can we deal with so many enquiries within as less time as possible? Because of the large number of data, the gap between CPU speed and the size of main memory has increasing considerably. One way to reduce the time to handle enquiries is to reduce the I/O number between the buffer and the secondary storage.An effective clustering of the objects can minimize the I/O cost between them. In this paper, according to the characteristic of the moving object database, we analyze the objects in buffer, according to their mappings in the two dimension coordinate, and then develop a density based clustering method to effectively reorganize the clusters. This new mechanism leads to the less cost of the I/O operation and the more efficient response to enquiries.
基金The author extends his appreciation to theDeputyship forResearch&Innovation,Ministry of Education in Saudi Arabia for funding this research work through the project number(IFPSAU-2021/01/17758).
文摘Finding clusters based on density represents a significant class of clustering algorithms.These methods can discover clusters of various shapes and sizes.The most studied algorithm in this class is theDensity-Based Spatial Clustering of Applications with Noise(DBSCAN).It identifies clusters by grouping the densely connected objects into one group and discarding the noise objects.It requires two input parameters:epsilon(fixed neighborhood radius)and MinPts(the lowest number of objects in epsilon).However,it can’t handle clusters of various densities since it uses a global value for epsilon.This article proposes an adaptation of the DBSCAN method so it can discover clusters of varied densities besides reducing the required number of input parameters to only one.Only user input in the proposed method is the MinPts.Epsilon on the other hand,is computed automatically based on statistical information of the dataset.The proposed method finds the core distance for each object in the dataset,takes the average of these distances as the first value of epsilon,and finds the clusters satisfying this density level.The remaining unclustered objects will be clustered using a new value of epsilon that equals the average core distances of unclustered objects.This process continues until all objects have been clustered or the remaining unclustered objects are less than 0.006 of the dataset’s size.The proposed method requires MinPts only as an input parameter because epsilon is computed from data.Benchmark datasets were used to evaluate the effectiveness of the proposed method that produced promising results.Practical experiments demonstrate that the outstanding ability of the proposed method to detect clusters of different densities even if there is no separation between them.The accuracy of the method ranges from 92%to 100%for the experimented datasets.
文摘We propose a new clustering algorithm that assists the researchers to quickly and accurately analyze data. We call this algorithm Combined Density-based and Constraint-based Algorithm (CDC). CDC consists of two phases. In the first phase, CDC employs the idea of density-based clustering algorithm to split the original data into a number of fragmented clusters. At the same time, CDC cuts off the noises and outliers. In the second phase, CDC employs the concept of K-means clustering algorithm to select a greater cluster to be the center. Then, the greater cluster merges some smaller clusters which satisfy some constraint rules. Due to the merged clusters around the center cluster, the clustering results show high accuracy. Moreover, CDC reduces the calculations and speeds up the clustering process. In this paper, the accuracy of CDC is evaluated and compared with those of K-means, hierarchical clustering, and the genetic clustering algorithm (GCA) proposed in 2004. Experimental results show that CDC has better performance.
基金supported by the National Natural Science Foundation of China(Nos.U1866602 and 71773025)the National Key Research and Development Program of China(No.2020YFB1006104)
文摘Density-based clustering is an important category among clustering algorithms. In real applications, many datasets suffer from incompleteness. Traditional imputation technologies or other techniques for handling missing values are not suitable for density-based clustering and decrease clustering result quality. To avoid these problems,we develop a novel density-based clustering approach for incomplete data based on Bayesian theory, which conducts imputation and clustering concurrently and makes use of intermediate clustering results. To avoid the impact of low-density areas inside non-convex clusters, we introduce a local imputation clustering algorithm, which aims to impute points to high-density local areas. The performances of the proposed algorithms are evaluated using ten synthetic datasets and five real-world datasets with induced missing values. The experimental results show the effectiveness of the proposed algorithms.
基金supported by the Career-FIT Fellowshipsfunded through European Union's Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No.713654supported by ACCORD(ITMS project code:313021X329),funded through the European Regional Development Fund.
文摘Ball milling is widely used in industry to mill particulate material.The primary purpose of this process is to attain an appropriate product size with the least possible energy consumption.The process is also extensively utilised in pharmaceuticals for the comminution of the excipients or drugs.Surprisingly,for ball mill,little is known concerning the mechanism of size reduction.Traditional prediction approaches are not deemed useful to provide significant insights into the operation or facilitate radical step changes in performance.Therefore,the discrete element method(DEM)as a computational modelling approach has been used in this paper.In previous research,DEM has been applied to simulate breaking behaviour through the impact energy of all ball collisions as the driving force for fracturing.However,the nature of pharmaceutical material fragmentation during ball milling is more complex.Suitable functional equations which link broken media and applied energy do not consider the collision of particulate media of different shapes or collisions of particulate media(such as granules)with balls and rotating mill drum.This could have a significant impact on fragmentation.Therefore,this paper aimed to investigate the fragmentation of bounded particles into DEM granules of different shape/size during the ball milling process.A systematic study was undertaken to explore the effect of milling speed on breakage behaviour.Also,in this study,a combination of a density-based clustering method and discrete element method was employed to numerically investigate the number and size of the fragments generated during the ball milling process over time.It was discovered that the collisions of the ball increased proportionally with rotation speed until reaching the critical rotation speed.Consequently,results illustrate that with an increase of rotation speed,the mill power increased correspondingly.The caratacting motion of mill material together with balls was identified as the most effective regime regarding the fragmentation,and fewer breakage events occurred for centrifugal motion.Higher quantities of the fines in each batch were produced with increased milling speed with less quantities of grain fragments.Moreover,the relationship between the number of produced fragment and milling speed at the end of the process exhibited a linear tendency.
基金supported by the National Natural Science Foundation of China(71271018)
文摘Overlapping community detection in a network is a challenging issue which attracts lots of attention in recent years.A notion of hesitant node（HN） is proposed. An HN contacts with multiple communities while the communications are not strong or even accidental, thus the HN holds an implicit community structure.However, HNs are not rare in the real world network. It is important to identify them because they can be efficient hubs which form the overlapping portions of communities or simple attached nodes to some communities. Current approaches have difficulties in identifying and clustering HNs. A density-based rough set model（DBRSM） is proposed by combining the virtue of densitybased algorithms and rough set models. It incorporates the macro perspective of the community structure of the whole network and the micro perspective of the local information held by HNs, which would facilitate the further ＂growth＂ of HNs in community. We offer a theoretical support for this model from the point of strength of the trust path. The experiments on the real-world and synthetic datasets show the practical significance of analyzing and clustering the HNs based on DBRSM. Besides, the clustering based on DBRSM promotes the modularity optimization.
文摘Under hydrothermal and solvothermal conditions,two novel cobalt-based complexes,{[Co_(2)(CIA)(OH)(1,4-dtb)]·3.2H_(2)O}n(HU23)and{[Co_(2)(CIA)(OH)(1,4-dib)]·3.5H2O·DMF}n(HU24),were successfully constructed by coordinatively assembling the semi-rigid multidentate ligand 5-(1-carboxyethoxy)isophthalic acid(H₃CIA)with the Nheterocyclic ligands 1,4-di(4H-1,2,4-triazol-4-yl)benzene(1,4-dtb)and 1,4-di(1H-imidazol-1-yl)benzene(1,4-dib),respectively,around Co^(2+)ions.Single-crystal X-ray diffraction analysis revealed that in both complexes HU23 and HU24,the CIA^(3-)anions adopt aκ^(7)-coordination mode,bridging six Co^(2+)ions via their five carboxylate oxygen atoms and one ether oxygen atom.This linkage forms tetranuclear[Co4(μ3-OH)2]^(6+)units.These Co-oxo cluster units were interconnected by CIA^(3-)anions to assemble into 2D kgd-type structures featuring a 3,6-connected topology.The 2D layers were further connected by 1,4-dtb and 1,4-dib,resulting in 3D pillar-layered frameworks for HU23 and HU24.Notably,despite the similar configurations of 1,4-dtb and 1,4-dib,differences in their coordination spatial orientations lead to topological divergence in the 3D frameworks of HU23 and HU24.Topological analysis indicates that the frameworks of HU23 and HU24 can be simplified into a 3,10-connected net(point symbol:(4^(10).6^(3).8^(2))(4^(3))_(2))and a 3,8-connected tfz-d net(point symbol:(4^(3))_(2)((4^(6).6^(18).8^(4)))),respectively.This structural differentiation confirms the precise regulatory role of ligands on the topology of metal-organic frameworks.Moreover,the ultraviolet-visible absorption spectra confirmed that HU23 and HU24 have strong absorption capabilities for ultraviolet and visible light.According to the Kubelka-Munk method,their bandwidths were 2.15 and 2.08 eV,respectively,which are consistent with those of typical semiconductor materials.Variable-temperature magnetic susceptibility measurements(2-300 K)revealed significant antiferromagnetic coupling in both complexes,with their effective magnetic moments decreasing markedly as the temperature lowered.CCDC:2457554,HU23;2457553,HU24.
基金supported by the National Natural Science Foundation of China (12473105 and 12473106)the central government guides local funds for science and technology development (YDZJSX2024D049)the Graduate Student Practice and Innovation Program of Shanxi Province (2024SJ313)
文摘As large-scale astronomical surveys,such as the Sloan Digital Sky Survey(SDSS)and the Large Sky Area Multi-Object Fiber Spectroscopic Telescope(LAMOST),generate increasingly complex datasets,clustering algorithms have become vital for identifying patterns and classifying celestial objects.This paper systematically investigates the application of five main categories of clustering techniques-partition-based,density-based,model-based,hierarchical,and“others”-across a range of astronomical research over the past decade.This review focuses on the six key application areas of stellar classification,galaxy structure analysis,detection of galactic and interstellar features,highenergy astrophysics,exoplanet studies,and anomaly detection.This paper provides an in-depth analysis of the performance and results of each method,considering their respective suitabilities for different data types.Additionally,it presents clustering algorithm selection strategies based on the characteristics of the spectroscopic data being analyzed.We highlight challenges such as handling large datasets,the need for more efficient computational tools,and the lack of labeled data.We also underscore the potential of unsupervised and semi-supervised clustering approaches to overcome these challenges,offering insight into their practical applications,performance,and results in astronomical research.
基金Under the auspices of National Natural Science Foundation of China(No.42571219)Key Project of Zhejiang Province Soft Science Research Plan(No.2023C25014)。
文摘Bottom-up and top-down endogenous automobile clusters exhibit distinct evolutionary traits and driving mechanisms,yet their comparative analysis remains understudied.Therefore,using Taizhou automobile industry cluster(TAIC)and Wuhu automobile industry cluster(WAIC)as cases,using historical statistical data and field interview data from the 1980s to 2023,combined with qualitative research methods of thematic and diachronic analysis,and quantitative research methods of social network analysis,we compare both endogenous automobile clusters’evolutionary traits and driving mechanisms.The results confirm both clusters undergo multi-scale spatial reconfiguration,organizational complexification,and intelligent networking technological transformation,yet diverge fundamentally:TAIC evolves through market-driven progressive expansion,transitioning from single to dual-core structures via private enterprise networking,with innovation following market-integrated logic and institutional thickness built on demand-driven evolution.Conversely,WAIC follows planned expansion,maintaining state-led hierarchical single-core stability through policy-driven breakthrough innovation and supply-dominated institutional construction-though both ultimately require formal-informal system synergy.Their coevolution is driven by dynamic interactions of path dependence(weakening influence),learning-innovation(strengthening influence),and relationship selection(inverted U-shaped trajectory),with divergent development paths rooted in TAIC’s grassroots self-organization genes versus WAIC’s top-level design genes,amplified by core enterprises’strategic disparities.The research findings can not only provide decision-making support for China’s industrial upgrading,but also contribute China’s insights to global economic governance.
基金supported by the Research Project of China Southern Power Grid(No.056200KK52222031).
文摘This paper proposes an equivalent modeling method for photovoltaic(PV)power stations via a particle swarm optimization(PSO)K-means clustering(KMC)algorithm with passive filter parameter clustering to address the complexities,simulation time cost and convergence problems of detailed PV power station models.First,the amplitude–frequency curves of different filter parameters are analyzed.Based on the results,a grouping parameter set for characterizing the external filter characteristics is established.These parameters are further defined as clustering parameters.A single PV inverter model is then established as a prerequisite foundation.The proposed equivalent method combines the global search capability of PSO with the rapid convergence of KMC,effectively overcoming the tendency of KMC to become trapped in local optima.This approach enhances both clustering accuracy and numerical stability when determining equivalence for PV inverter units.Using the proposed clustering method,both a detailed PV power station model and an equivalent model are developed and compared.Simulation and hardwarein-loop(HIL)results based on the equivalent model verify that the equivalent method accurately represents the dynamic characteristics of PVpower stations and adapts well to different operating conditions.The proposed equivalent modeling method provides an effective analysis tool for future renewable energy integration research.
文摘Villager Pan Chunlin is witnessing a boom in his homestay business.More and more visitors are coming to his village,Yucun Village in Anji County,Huzhou City,Zhejiang Province.
基金Under the auspices of the Key Projects of Philosophy and Social Sciences Research,Ministry of Education of China(No.23JZD008)National Natural Science Foundation of China(No.42171193)+2 种基金Key Project of Guangdong Provincial Philosophy and Social Sciences Planning(No.GD24ES013,GD25ZX04)2025 Guangzhou Basic and Applied Basic Research Special Project(No.2025A04J7127)Fundamental Research Funds for the Central Universities,Sun Yat-sen University(No.24wkjc11)。
文摘The shift toward specialized and large-scale agricultural production has spurred the emergence of agricultural clusters as key forces of rural vitalization and sustainable development.This paper explored the formation and evolution of Meizhou pomelo industry cluster in China,focusing on its role in restructuring rural socio-economic systems and integrating the whole value chains.Based on a case study employing qualitative methods such as in-depth interviews and participatory observation,the agricultural cluster evolution of Meizhou pomelo was categorized into three key phases of initial decentralization,self-organized scaling,and reorganized clustering.Geographical proximity and industrial agglomeration constitute the physical foundation,while vertical/horizontal linkages,technologic-al innovation,and policy support enhance competitiveness.Special mechanisms emerge through localized social networks,farmer co-operatives’activation,and cross-regional market expansion.The cluster’s impact is manifested in the shift from extensive to standard-ized and modernized production,diversified and flexible livelihood of farmers,and the integration of agriculture with industry and ser-vices.The development of the whole value chain based on agricultural cluster represents a critical pathway for achieving agricultural modernization,encompassing both internal and external value chain optimization.Through quality assurance systems,product diversi-fication strategies,operational efficiency improvements,and brand enhancement,these clusters amplify product value propositions and market competitiveness.This systemic approach facilitates supply-demand coordination,enables resource synergies,and optimizes eco-nomic returns across the horizontal and vertical value chain.This paper argues that agricultural clusters serve as strategic catalysts for sustainable rural development by reconstructing local production systems,fostering innovation ecosystems,and aligning agricultural modernization.It contributes to debates on rural vitalization by demonstrating how agricultural clustering can reconfigure rural areas as hubs of ecological modernization,rather than mere urban peripheries.
文摘By encouraging collaboration,attracting young entrepreneurs,and setting up village-invested enterprises,rural communities in Zhejiang Province are growing their income streams.“Over the past years,we have seen more and more tourists coming to our village,and their stay here has grown longer.Many even said they don’t want to leave,”Pan Chunlin,a resident of Yucun Village in Anji County of Huzhou City,east China’s Zhejiang Province,told China Today.Pan is the owner of the Chunlin Lodge,a bed&breakfast(B&B)which received more than 70,000 visitors last year,generating over RMB 4 million in revenue.
文摘Noncohesive particle clusters are identified and tracked in turbulent flows to determine the breakdown and time evolution of cluster statistics and their implications for interscale mass transfer,which has connections to the classical turbulent energy cascade and its mass cascade counterpart running in parallel.In particular,the formation and dynamics of sediment and larvae clusters are of interest to coral larvae settlement in coastal regions and particularly the resilience of green-gray coastal protection solutions.Analogous cluster behavior is relevant to cloud microphysics and precipitation initiation,radiation transport and light transmission through colloids and suspensions,heat and mass transfer in particle-laden flows,and viral and pollutant transmission.Following a comparison between various clustering techniques,we adopt a density-based cluster identification algorithm based on its simplicity and efficiency,where particles are clustered based on the number of neighboring particles in their individual spheres of influence.We establish parallels with lattice-based percolation theory,as evident in the power-law scaling of the cluster size distribution near the percolation threshold.The degree of discontinuity of the phase transition associated with this percolation threshold is observed to broaden with larger Stokes numbers and thereby large-scale clustering.The sensitivity of our findings to the employed clustering algorithm is discussed.A novel cluster tracking algorithm is deployed to determine the interscale transfer rate along the particle-number phase-space dimension via accounting of cluster breakup and merger events,extending previous work on the bubble breakup cascade beneath surface breaking waves.Our findings shed light on the interaction between particle clusters and their carrier turbulent flows,with an eye toward transport models incorporating cluster characteristics and dynamics.
基金Supported by the Korea Health Technology R&D Project through the Korea Health Industry Development Institute(KHIDI),the Ministry of Health&Welfare,Republic of Korea(No.RS-2020-KH088726)the Patient-Centered Clinical Research Coordinating Center(PACEN),the Ministry of Health and Welfare,Republic of Korea(No.HC19C0276)the National Research Foundation of Korea(NRF),the Korea Government(MSIT)(No.RS-2023-00247504).
文摘AIM:To evaluate long-term visual field(VF)prediction using K-means clustering in patients with primary open angle glaucoma(POAG).METHODS:Patients who underwent 24-2 VF tests≥10 were included in this study.Using 52 total deviation values(TDVs)from the first 10 VF tests of the training dataset,VF points were clustered into several regions using the hierarchical ordered partitioning and collapsing hybrid(HOPACH)and K-means clustering.Based on the clustering results,a linear regression analysis was applied to each clustered region of the testing dataset to predict the TDVs of the 10th VF test.Three to nine VF tests were used to predict the 10th VF test,and the prediction errors(root mean square error,RMSE)of each clustering method and pointwise linear regression(PLR)were compared.RESULTS:The training group consisted of 228 patients(mean age,54.20±14.38y;123 males and 105 females),and the testing group included 81 patients(mean age,54.88±15.22y;43 males and 38 females).All subjects were diagnosed with POAG.Fifty-two VF points were clustered into 11 and nine regions using HOPACH and K-means clustering,respectively.K-means clustering had a lower prediction error than PLR when n=1:3 and 1:4(both P≤0.003).The prediction errors of K-means clustering were lower than those of HOPACH in all sections(n=1:4 to 1:9;all P≤0.011),except for n=1:3(P=0.680).PLR outperformed K-means clustering only when n=1:8 and 1:9(both P≤0.020).CONCLUSION:K-means clustering can predict longterm VF test results more accurately in patients with POAG with limited VF data.
