Many fields,such as neuroscience,are experiencing the vast prolife ration of cellular data,underscoring the need fo r organizing and interpreting large datasets.A popular approach partitions data into manageable subse...Many fields,such as neuroscience,are experiencing the vast prolife ration of cellular data,underscoring the need fo r organizing and interpreting large datasets.A popular approach partitions data into manageable subsets via hierarchical clustering,but objective methods to determine the appropriate classification granularity are missing.We recently introduced a technique to systematically identify when to stop subdividing clusters based on the fundamental principle that cells must differ more between than within clusters.Here we present the corresponding protocol to classify cellular datasets by combining datadriven unsupervised hierarchical clustering with statistical testing.These general-purpose functions are applicable to any cellular dataset that can be organized as two-dimensional matrices of numerical values,including molecula r,physiological,and anatomical datasets.We demonstrate the protocol using cellular data from the Janelia MouseLight project to chara cterize morphological aspects of neurons.展开更多
Customer segmentation according to load-shape profiles using smart meter data is an increasingly important application to vital the planning and operation of energy systems and to enable citizens’participation in the...Customer segmentation according to load-shape profiles using smart meter data is an increasingly important application to vital the planning and operation of energy systems and to enable citizens’participation in the energy transition.This study proposes an innovative multi-step clustering procedure to segment customers based on load-shape patterns at the daily and intra-daily time horizons.Smart meter data is split between daily and hourly normalized time series to assess monthly,weekly,daily,and hourly seasonality patterns separately.The dimensionality reduction implicit in the splitting allows a direct approach to clustering raw daily energy time series data.The intraday clustering procedure sequentially identifies representative hourly day-unit profiles for each customer and the entire population.For the first time,a step function approach is applied to reduce time series dimensionality.Customer attributes embedded in surveys are employed to build external clustering validation metrics using Cramer’s V correlation factors and to identify statistically significant determinants of load-shape in energy usage.In addition,a time series features engineering approach is used to extract 16 relevant demand flexibility indicators that characterize customers and corresponding clusters along four different axes:available Energy(E),Temporal patterns(T),Consistency(C),and Variability(V).The methodology is implemented on a real-world electricity consumption dataset of 325 Small and Medium-sized Enterprise(SME)customers,identifying 4 daily and 6 hourly easy-to-interpret,well-defined clusters.The application of the methodology includes selecting key parameters via grid search and a thorough comparison of clustering distances and methods to ensure the robustness of the results.Further research can test the scalability of the methodology to larger datasets from various customer segments(households and large commercial)and locations with different weather and socioeconomic conditions.展开更多
Domaining is a crucial process in geostatistics, particularly when significant spatial variations are observed within a site, as these variations can significantly affect the outcomes of spatial modeling. This study i...Domaining is a crucial process in geostatistics, particularly when significant spatial variations are observed within a site, as these variations can significantly affect the outcomes of spatial modeling. This study investigates the application of hard and fuzzy clustering algorithms for domain delineation, using geological and geochemical data from two exploration campaigns at the eastern Kahang deposit in central Iran. The dataset includes geological layers (lithology, alteration, and mineral zones), geochemical layers (Cu, Mo, Ag, and Au grades), and borehole coordinates. Six clustering algorithms—K-means, hierarchical, affinity propagation, self-organizing map (SOM), fuzzy C-means, and Gustafson-Kessel—were applied to determine the optimal number of clusters, which ranged from 3 to 4. The fuzziness and weighting parameters were found to range from 1.1 to 1.3 and 0.1 to 0.3, respectively, based on the evaluation of various hard and fuzzy cluster validity indices. Directional variograms were computed to assess spatial anisotropy, and the anisotropy ellipsoid for each domain was defined to identify the model with the highest level of anisotropic discrimination among the domains. The SOM algorithm, which incorporated both qualitative and quantitative data, produced the best model, resulting in the identification of three distinct domains. These findings underscore the effectiveness of combining clustering techniques with variogram analysis for accurate domain delineation in geostatistical modeling.展开更多
Existing multi-view deep subspace clustering methods aim to learn a unified representation from multi-view data,while the learned representation is difficult to maintain the underlying structure hidden in the origin s...Existing multi-view deep subspace clustering methods aim to learn a unified representation from multi-view data,while the learned representation is difficult to maintain the underlying structure hidden in the origin samples,especially the high-order neighbor relationship between samples.To overcome the above challenges,this paper proposes a novel multi-order neighborhood fusion based multi-view deep subspace clustering model.We creatively integrate the multi-order proximity graph structures of different views into the self-expressive layer by a multi-order neighborhood fusion module.By this design,the multi-order Laplacian matrix supervises the learning of the view-consistent self-representation affinity matrix;then,we can obtain an optimal global affinity matrix where each connected node belongs to one cluster.In addition,the discriminative constraint between views is designed to further improve the clustering performance.A range of experiments on six public datasets demonstrates that the method performs better than other advanced multi-view clustering methods.The code is available at https://github.com/songzuolong/MNF-MDSC(accessed on 25 December 2024).展开更多
The characterization and clustering of rock discontinuity sets are a crucial and challenging task in rock mechanics and geotechnical engineering.Over the past few decades,the clustering of discontinuity sets has under...The characterization and clustering of rock discontinuity sets are a crucial and challenging task in rock mechanics and geotechnical engineering.Over the past few decades,the clustering of discontinuity sets has undergone rapid and remarkable development.However,there is no relevant literature summarizing these achievements,and this paper attempts to elaborate on the current status and prospects in this field.Specifically,this review aims to discuss the development process of clustering methods for discontinuity sets and the state-of-the-art relevant algorithms.First,we introduce the importance of discontinuity clustering analysis and follow the comprehensive characterization approaches of discontinuity data.A bibliometric analysis is subsequently conducted to clarify the current status and development characteristics of the clustering of discontinuity sets.The methods for the clustering analysis of rock discontinuities are reviewed in terms of single-and multi-parameter clustering methods.Single-parameter methods can be classified into empirical judgment methods,dynamic clustering methods,relative static clustering methods,and static clustering methods,reflecting the continuous optimization and improvement of clustering algorithms.Moreover,this paper compares the current mainstream of single-parameter clustering methods with multi-parameter clustering methods.It is emphasized that the current single-parameter clustering methods have reached their performance limits,with little room for improvement,and that there is a need to extend the study of multi-parameter clustering methods.Finally,several suggestions are offered for future research on the clustering of discontinuity sets.展开更多
European directives advocate for end-users to be aware of their energy consumption.However,individual energy monitoring tools,such as energy meters or cost allocators,are not always affordable or technically feasible ...European directives advocate for end-users to be aware of their energy consumption.However,individual energy monitoring tools,such as energy meters or cost allocators,are not always affordable or technically feasible to install.Therefore,the development of virtual tools that enable the study of energy consumption in existing buildings is necessary.Virtual sensors,particularly based on white-box models,offer the opportunity to recreate these behaviours.When calibrated with measured data,white-box models,which incorporate detailed building physics,become increasingly valuable for designing energy-efficient buildings.This research explores a novel approach to identifying building’s load period directly from energy data generated by these calibrated models.The volume of data generated by white-box models can be overwhelming for visual analysis,but the hypothesis here is that analysing this data through clustering techniques can reveal patterns related to occupant behaviour and operational schedules.By feeding indoor temperature data into the calibrated model and analysing the resulting energy outputs,the research proposes a method to identify the heating,ventilation and air conditioning(HVAC)system operation schedule,free oscillation periods and non-recurrent events.Validation is achieved by comparing the identified periods with actual measured data.This methodology enables the development of a virtual sensor for cost allocation,which minimises the need for physical sensor deployment while complying with European Union directives.The research not only demonstrates high accuracy but also the potential to outperform measured schedule.This suggests the ability of the method to identify missing sensor data or other factors affecting temperature curves,enabling fault detection and diagnostics(FDD).Consequently,this opens doors for setting optimised operation schedules that balance energy efficiency with occupant comfort.展开更多
Reliable Cluster Head(CH)selectionbased routing protocols are necessary for increasing the packet transmission efficiency with optimal path discovery that never introduces degradation over the transmission reliability...Reliable Cluster Head(CH)selectionbased routing protocols are necessary for increasing the packet transmission efficiency with optimal path discovery that never introduces degradation over the transmission reliability.In this paper,Hybrid Golden Jackal,and Improved Whale Optimization Algorithm(HGJIWOA)is proposed as an effective and optimal routing protocol that guarantees efficient routing of data packets in the established between the CHs and the movable sink.This HGJIWOA included the phases of Dynamic Lens-Imaging Learning Strategy and Novel Update Rules for determining the reliable route essential for data packets broadcasting attained through fitness measure estimation-based CH selection.The process of CH selection achieved using Golden Jackal Optimization Algorithm(GJOA)completely depends on the factors of maintainability,consistency,trust,delay,and energy.The adopted GJOA algorithm play a dominant role in determining the optimal path of routing depending on the parameter of reduced delay and minimal distance.It further utilized Improved Whale Optimisation Algorithm(IWOA)for forwarding the data from chosen CHs to the BS via optimized route depending on the parameters of energy and distance.It also included a reliable route maintenance process that aids in deciding the selected route through which data need to be transmitted or re-routed.The simulation outcomes of the proposed HGJIWOA mechanism with different sensor nodes confirmed an improved mean throughput of 18.21%,sustained residual energy of 19.64%with minimized end-to-end delay of 21.82%,better than the competitive CH selection approaches.展开更多
Music recommendation systems are essential due to the vast amount of music available on streaming platforms,which can overwhelm users trying to find new tracks that match their preferences.These systems analyze users...Music recommendation systems are essential due to the vast amount of music available on streaming platforms,which can overwhelm users trying to find new tracks that match their preferences.These systems analyze users’emotional responses,listening habits,and personal preferences to provide personalized suggestions.A significant challenge they face is the“cold start”problem,where new users have no past interactions to guide recommendations.To improve user experience,these systems aimto effectively recommendmusic even to such users by considering their listening behavior and music popularity.This paper introduces a novel music recommendation system that combines order clustering and a convolutional neural network,utilizing user comments and rankings as input.Initially,the system organizes users into clusters based on semantic similarity,followed by the utilization of their rating similarities as input for the convolutional neural network.This network then predicts ratings for unreviewed music by users.Additionally,the system analyses user music listening behaviour and music popularity.Music popularity can help to address cold start users as well.Finally,the proposed method recommends unreviewed music based on predicted high rankings and popularity,taking into account each user’s music listening habits.The proposed method combines predicted high rankings and popularity by first selecting popular unreviewedmusic that themodel predicts to have the highest ratings for each user.Among these,the most popular tracks are prioritized,defined by metrics such as frequency of listening across users.The number of recommended tracks is aligned with each user’s typical listening rate.The experimental findings demonstrate that the new method outperformed other classification techniques and prior recommendation systems,yielding a mean absolute error(MAE)rate and rootmean square error(RMSE)rate of approximately 0.0017,a hit rate of 82.45%,an average normalized discounted cumulative gain(nDCG)of 82.3%,and a prediction accuracy of new ratings at 99.388%.展开更多
In order to solve the problems of short network lifetime and high data transmission delay in data gathering for wireless sensor network(WSN)caused by uneven energy consumption among nodes,a hybrid energy efficient clu...In order to solve the problems of short network lifetime and high data transmission delay in data gathering for wireless sensor network(WSN)caused by uneven energy consumption among nodes,a hybrid energy efficient clustering routing base on firefly and pigeon-inspired algorithm(FF-PIA)is proposed to optimise the data transmission path.After having obtained the optimal number of cluster head node(CH),its result might be taken as the basis of producing the initial population of FF-PIA algorithm.The L′evy flight mechanism and adaptive inertia weighting are employed in the algorithm iteration to balance the contradiction between the global search and the local search.Moreover,a Gaussian perturbation strategy is applied to update the optimal solution,ensuring the algorithm can jump out of the local optimal solution.And,in the WSN data gathering,a onedimensional signal reconstruction algorithm model is developed by dilated convolution and residual neural networks(DCRNN).We conducted experiments on the National Oceanic and Atmospheric Administration(NOAA)dataset.It shows that the DCRNN modeldriven data reconstruction algorithm improves the reconstruction accuracy as well as the reconstruction time performance.FF-PIA and DCRNN clustering routing co-simulation reveals that the proposed algorithm can effectively improve the performance in extending the network lifetime and reducing data transmission delay.展开更多
To guarantee safe and efficient tunneling of a tunnel boring machine(TBM),rapid and accurate judgment of the rock mass condition is essential.Based on fuzzy C-means clustering,this paper proposes a grouped machine lea...To guarantee safe and efficient tunneling of a tunnel boring machine(TBM),rapid and accurate judgment of the rock mass condition is essential.Based on fuzzy C-means clustering,this paper proposes a grouped machine learning method for predicting rock mass parameters.An elaborate data set on field rock mass is collected,which also matches field TBM tunneling.Meanwhile,target stratum samples are divided into several clusters by fuzzy C-means clustering,and multiple submodels are trained by samples in different clusters with the input of pretreated TBM tunneling data and the output of rock mass parameter data.Each testing sample or newly encountered tunneling condition can be predicted by multiple submodels with the weight of the membership degree of the sample to each cluster.The proposed method has been realized by 100 training samples and verified by 30 testing samples collected from the C1 part of the Pearl Delta water resources allocation project.The average percentage error of uniaxial compressive strength and joint frequency(Jf)of the 30 testing samples predicted by the pure back propagation(BP)neural network is 13.62%and 12.38%,while that predicted by the BP neural network combined with fuzzy C-means is 7.66%and6.40%,respectively.In addition,by combining fuzzy C-means clustering,the prediction accuracies of support vector regression and random forest are also improved to different degrees,which demonstrates that fuzzy C-means clustering is helpful for improving the prediction accuracy of machine learning and thus has good applicability.Accordingly,the proposed method is valuable for predicting rock mass parameters during TBM tunneling.展开更多
In recent years,many unknown protocols are constantly emerging,and they bring severe challenges to network security and network management.Existing unknown protocol recognition methods suffer from weak feature extract...In recent years,many unknown protocols are constantly emerging,and they bring severe challenges to network security and network management.Existing unknown protocol recognition methods suffer from weak feature extraction ability,and they cannot mine the discriminating features of the protocol data thoroughly.To address the issue,we propose an unknown application layer protocol recognition method based on deep clustering.Deep clustering which consists of the deep neural network and the clustering algorithm can automatically extract the features of the input and cluster the data based on the extracted features.Compared with the traditional clustering methods,deep clustering boasts of higher clustering accuracy.The proposed method utilizes network-in-network(NIN),channel attention,spatial attention and Bidirectional Long Short-term memory(BLSTM)to construct an autoencoder to extract the spatial-temporal features of the protocol data,and utilizes the unsupervised clustering algorithm to recognize the unknown protocols based on the features.The method firstly extracts the application layer protocol data from the network traffic and transforms the data into one-dimensional matrix.Secondly,the autoencoder is pretrained,and the protocol data is compressed into low dimensional latent space by the autoencoder and the initial clustering is performed with K-Means.Finally,the clustering loss is calculated and the classification model is optimized according to the clustering loss.The classification results can be obtained when the classification model is optimal.Compared with the existing unknown protocol recognition methods,the proposed method utilizes deep clustering to cluster the unknown protocols,and it can mine the key features of the protocol data and recognize the unknown protocols accurately.Experimental results show that the proposed method can effectively recognize the unknown protocols,and its performance is better than other methods.展开更多
The study delves into the expanding role of network platforms in our daily lives, encompassing various mediums like blogs, forums, online chats, and prominent social media platforms such as Facebook, Twitter, and Inst...The study delves into the expanding role of network platforms in our daily lives, encompassing various mediums like blogs, forums, online chats, and prominent social media platforms such as Facebook, Twitter, and Instagram. While these platforms offer avenues for self-expression and community support, they concurrently harbor negative impacts, fostering antisocial behaviors like phishing, impersonation, hate speech, cyberbullying, cyberstalking, cyberterrorism, fake news propagation, spamming, and fraud. Notably, individuals also leverage these platforms to connect with authorities and seek aid during disasters. The overarching objective of this research is to address the dual nature of network platforms by proposing innovative methodologies aimed at enhancing their positive aspects and mitigating their negative repercussions. To achieve this, the study introduces a weight learning method grounded in multi-linear attribute ranking. This approach serves to evaluate the significance of attribute combinations across all feature spaces. Additionally, a novel clustering method based on tensors is proposed to elevate the quality of clustering while effectively distinguishing selected features. The methodology incorporates a weighted average similarity matrix and optionally integrates weighted Euclidean distance, contributing to a more nuanced understanding of attribute importance. The analysis of the proposed methods yields significant findings. The weight learning method proves instrumental in discerning the importance of attribute combinations, shedding light on key aspects within feature spaces. Simultaneously, the clustering method based on tensors exhibits improved efficacy in enhancing clustering quality and feature distinction. This not only advances our understanding of attribute importance but also paves the way for more nuanced data analysis methodologies. In conclusion, this research underscores the pivotal role of network platforms in contemporary society, emphasizing their potential for both positive contributions and adverse consequences. The proposed methodologies offer novel approaches to address these dualities, providing a foundation for future research and practical applications. Ultimately, this study contributes to the ongoing discourse on optimizing the utility of network platforms while minimizing their negative impacts.展开更多
Although many multi-view clustering(MVC) algorithms with acceptable performances have been presented, to the best of our knowledge, nearly all of them need to be fed with the correct number of clusters. In addition, t...Although many multi-view clustering(MVC) algorithms with acceptable performances have been presented, to the best of our knowledge, nearly all of them need to be fed with the correct number of clusters. In addition, these existing algorithms create only the hard and fuzzy partitions for multi-view objects,which are often located in highly-overlapping areas of multi-view feature space. The adoption of hard and fuzzy partition ignores the ambiguity and uncertainty in the assignment of objects, likely leading to performance degradation. To address these issues, we propose a novel sparse reconstructive multi-view evidential clustering algorithm(SRMVEC). Based on a sparse reconstructive procedure, SRMVEC learns a shared affinity matrix across views, and maps multi-view objects to a 2-dimensional humanreadable chart by calculating 2 newly defined mathematical metrics for each object. From this chart, users can detect the number of clusters and select several objects existing in the dataset as cluster centers. Then, SRMVEC derives a credal partition under the framework of evidence theory, improving the fault tolerance of clustering. Ablation studies show the benefits of adopting the sparse reconstructive procedure and evidence theory. Besides,SRMVEC delivers effectiveness on benchmark datasets by outperforming some state-of-the-art methods.展开更多
Multi-view Subspace Clustering (MVSC) emerges as an advanced clustering method, designed to integrate diverse views to uncover a common subspace, enhancing the accuracy and robustness of clustering results. The signif...Multi-view Subspace Clustering (MVSC) emerges as an advanced clustering method, designed to integrate diverse views to uncover a common subspace, enhancing the accuracy and robustness of clustering results. The significance of low-rank prior in MVSC is emphasized, highlighting its role in capturing the global data structure across views for improved performance. However, it faces challenges with outlier sensitivity due to its reliance on the Frobenius norm for error measurement. Addressing this, our paper proposes a Low-Rank Multi-view Subspace Clustering Based on Sparse Regularization (LMVSC- Sparse) approach. Sparse regularization helps in selecting the most relevant features or views for clustering while ignoring irrelevant or noisy ones. This leads to a more efficient and effective representation of the data, improving the clustering accuracy and robustness, especially in the presence of outliers or noisy data. By incorporating sparse regularization, LMVSC-Sparse can effectively handle outlier sensitivity, which is a common challenge in traditional MVSC methods relying solely on low-rank priors. Then Alternating Direction Method of Multipliers (ADMM) algorithm is employed to solve the proposed optimization problems. Our comprehensive experiments demonstrate the efficiency and effectiveness of LMVSC-Sparse, offering a robust alternative to traditional MVSC methods.展开更多
In clustering algorithms,the selection of neighbors significantly affects the quality of the final clustering results.While various neighbor relationships exist,such as K-nearest neighbors,natural neighbors,and shared...In clustering algorithms,the selection of neighbors significantly affects the quality of the final clustering results.While various neighbor relationships exist,such as K-nearest neighbors,natural neighbors,and shared neighbors,most neighbor relationships can only handle single structural relationships,and the identification accuracy is low for datasets with multiple structures.In life,people’s first instinct for complex things is to divide them into multiple parts to complete.Partitioning the dataset into more sub-graphs is a good idea approach to identifying complex structures.Taking inspiration from this,we propose a novel neighbor method:Shared Natural Neighbors(SNaN).To demonstrate the superiority of this neighbor method,we propose a shared natural neighbors-based hierarchical clustering algorithm for discovering arbitrary-shaped clusters(HC-SNaN).Our algorithm excels in identifying both spherical clusters and manifold clusters.Tested on synthetic datasets and real-world datasets,HC-SNaN demonstrates significant advantages over existing clustering algorithms,particularly when dealing with datasets containing arbitrary shapes.展开更多
In this paper,we introduce a novel Multi-scale and Auto-tuned Semi-supervised Deep Subspace Clustering(MAS-DSC)algorithm,aimed at addressing the challenges of deep subspace clustering in high-dimensional real-world da...In this paper,we introduce a novel Multi-scale and Auto-tuned Semi-supervised Deep Subspace Clustering(MAS-DSC)algorithm,aimed at addressing the challenges of deep subspace clustering in high-dimensional real-world data,particularly in the field of medical imaging.Traditional deep subspace clustering algorithms,which are mostly unsupervised,are limited in their ability to effectively utilize the inherent prior knowledge in medical images.Our MAS-DSC algorithm incorporates a semi-supervised learning framework that uses a small amount of labeled data to guide the clustering process,thereby enhancing the discriminative power of the feature representations.Additionally,the multi-scale feature extraction mechanism is designed to adapt to the complexity of medical imaging data,resulting in more accurate clustering performance.To address the difficulty of hyperparameter selection in deep subspace clustering,this paper employs a Bayesian optimization algorithm for adaptive tuning of hyperparameters related to subspace clustering,prior knowledge constraints,and model loss weights.Extensive experiments on standard clustering datasets,including ORL,Coil20,and Coil100,validate the effectiveness of the MAS-DSC algorithm.The results show that with its multi-scale network structure and Bayesian hyperparameter optimization,MAS-DSC achieves excellent clustering results on these datasets.Furthermore,tests on a brain tumor dataset demonstrate the robustness of the algorithm and its ability to leverage prior knowledge for efficient feature extraction and enhanced clustering performance within a semi-supervised learning framework.展开更多
Traditional Fuzzy C-Means(FCM)and Possibilistic C-Means(PCM)clustering algorithms are data-driven,and their objective function minimization process is based on the available numeric data.Recently,knowledge hints have ...Traditional Fuzzy C-Means(FCM)and Possibilistic C-Means(PCM)clustering algorithms are data-driven,and their objective function minimization process is based on the available numeric data.Recently,knowledge hints have been introduced to formknowledge-driven clustering algorithms,which reveal a data structure that considers not only the relationships between data but also the compatibility with knowledge hints.However,these algorithms cannot produce the optimal number of clusters by the clustering algorithm itself;they require the assistance of evaluation indices.Moreover,knowledge hints are usually used as part of the data structure(directly replacing some clustering centers),which severely limits the flexibility of the algorithm and can lead to knowledgemisguidance.To solve this problem,this study designs a newknowledge-driven clustering algorithmcalled the PCM clusteringwith High-density Points(HP-PCM),in which domain knowledge is represented in the form of so-called high-density points.First,a newdatadensitycalculation function is proposed.The Density Knowledge Points Extraction(DKPE)method is established to filter out high-density points from the dataset to form knowledge hints.Then,these hints are incorporated into the PCM objective function so that the clustering algorithm is guided by high-density points to discover the natural data structure.Finally,the initial number of clusters is set to be greater than the true one based on the number of knowledge hints.Then,the HP-PCM algorithm automatically determines the final number of clusters during the clustering process by considering the cluster elimination mechanism.Through experimental studies,including some comparative analyses,the results highlight the effectiveness of the proposed algorithm,such as the increased success rate in clustering,the ability to determine the optimal cluster number,and the faster convergence speed.展开更多
Path-based clustering algorithms typically generate clusters by optimizing a benchmark function.Most optimiza-tion methods in clustering algorithms often offer solutions close to the general optimal value.This study a...Path-based clustering algorithms typically generate clusters by optimizing a benchmark function.Most optimiza-tion methods in clustering algorithms often offer solutions close to the general optimal value.This study achieves the global optimum value for the criterion function in a shorter time using the minimax distance,Maximum Spanning Tree“MST”,and meta-heuristic algorithms,including Genetic Algorithm“GA”and Particle Swarm Optimization“PSO”.The Fast Path-based Clustering“FPC”algorithm proposed in this paper can find cluster centers correctly in most datasets and quickly perform clustering operations.The FPC does this operation using MST,the minimax distance,and a new hybrid meta-heuristic algorithm in a few rounds of algorithm iterations.This algorithm can achieve the global optimal value,and the main clustering process of the algorithm has a computational complexity of O�k2×n�.However,due to the complexity of the minimum distance algorithm,the total computational complexity is O�n2�.Experimental results of FPC on synthetic datasets with arbitrary shapes demonstrate that the algorithm is resistant to noise and outliers and can correctly identify clusters of varying sizes and numbers.In addition,the FPC requires the number of clusters as the only parameter to perform the clustering process.A comparative analysis of FPC and other clustering algorithms in this domain indicates that FPC exhibits superior speed,stability,and performance.展开更多
Contrastive learning is a significant research direction in the field of deep learning.However,existing data augmentation methods often lead to issues such as semantic drift in generated views while the complexity of ...Contrastive learning is a significant research direction in the field of deep learning.However,existing data augmentation methods often lead to issues such as semantic drift in generated views while the complexity of model pre-training limits further improvement in the performance of existing methods.To address these challenges,we propose the Efficient Clustering Network based on Matrix Factorization(ECN-MF).Specifically,we design a batched low-rank Singular Value Decomposition(SVD)algorithm for data augmentation to eliminate redundant information and uncover major patterns of variation and key information in the data.Additionally,we design a Mutual Information-Enhanced Clustering Module(MI-ECM)to accelerate the training process by leveraging a simple architecture to bring samples from the same cluster closer while pushing samples from other clusters apart.Extensive experiments on six datasets demonstrate that ECN-MF exhibits more effective performance compared to state-of-the-art algorithms.展开更多
基金supported in part by NIH grants R01NS39600,U01MH114829RF1MH128693(to GAA)。
文摘Many fields,such as neuroscience,are experiencing the vast prolife ration of cellular data,underscoring the need fo r organizing and interpreting large datasets.A popular approach partitions data into manageable subsets via hierarchical clustering,but objective methods to determine the appropriate classification granularity are missing.We recently introduced a technique to systematically identify when to stop subdividing clusters based on the fundamental principle that cells must differ more between than within clusters.Here we present the corresponding protocol to classify cellular datasets by combining datadriven unsupervised hierarchical clustering with statistical testing.These general-purpose functions are applicable to any cellular dataset that can be organized as two-dimensional matrices of numerical values,including molecula r,physiological,and anatomical datasets.We demonstrate the protocol using cellular data from the Janelia MouseLight project to chara cterize morphological aspects of neurons.
基金supported by the Spanish Ministry of Science and Innovation under Projects PID2022-137680OB-C32 and PID2022-139187OB-I00.
文摘Customer segmentation according to load-shape profiles using smart meter data is an increasingly important application to vital the planning and operation of energy systems and to enable citizens’participation in the energy transition.This study proposes an innovative multi-step clustering procedure to segment customers based on load-shape patterns at the daily and intra-daily time horizons.Smart meter data is split between daily and hourly normalized time series to assess monthly,weekly,daily,and hourly seasonality patterns separately.The dimensionality reduction implicit in the splitting allows a direct approach to clustering raw daily energy time series data.The intraday clustering procedure sequentially identifies representative hourly day-unit profiles for each customer and the entire population.For the first time,a step function approach is applied to reduce time series dimensionality.Customer attributes embedded in surveys are employed to build external clustering validation metrics using Cramer’s V correlation factors and to identify statistically significant determinants of load-shape in energy usage.In addition,a time series features engineering approach is used to extract 16 relevant demand flexibility indicators that characterize customers and corresponding clusters along four different axes:available Energy(E),Temporal patterns(T),Consistency(C),and Variability(V).The methodology is implemented on a real-world electricity consumption dataset of 325 Small and Medium-sized Enterprise(SME)customers,identifying 4 daily and 6 hourly easy-to-interpret,well-defined clusters.The application of the methodology includes selecting key parameters via grid search and a thorough comparison of clustering distances and methods to ensure the robustness of the results.Further research can test the scalability of the methodology to larger datasets from various customer segments(households and large commercial)and locations with different weather and socioeconomic conditions.
文摘Domaining is a crucial process in geostatistics, particularly when significant spatial variations are observed within a site, as these variations can significantly affect the outcomes of spatial modeling. This study investigates the application of hard and fuzzy clustering algorithms for domain delineation, using geological and geochemical data from two exploration campaigns at the eastern Kahang deposit in central Iran. The dataset includes geological layers (lithology, alteration, and mineral zones), geochemical layers (Cu, Mo, Ag, and Au grades), and borehole coordinates. Six clustering algorithms—K-means, hierarchical, affinity propagation, self-organizing map (SOM), fuzzy C-means, and Gustafson-Kessel—were applied to determine the optimal number of clusters, which ranged from 3 to 4. The fuzziness and weighting parameters were found to range from 1.1 to 1.3 and 0.1 to 0.3, respectively, based on the evaluation of various hard and fuzzy cluster validity indices. Directional variograms were computed to assess spatial anisotropy, and the anisotropy ellipsoid for each domain was defined to identify the model with the highest level of anisotropic discrimination among the domains. The SOM algorithm, which incorporated both qualitative and quantitative data, produced the best model, resulting in the identification of three distinct domains. These findings underscore the effectiveness of combining clustering techniques with variogram analysis for accurate domain delineation in geostatistical modeling.
基金supported by the National Key R&D Program of China(2023YFC3304600).
文摘Existing multi-view deep subspace clustering methods aim to learn a unified representation from multi-view data,while the learned representation is difficult to maintain the underlying structure hidden in the origin samples,especially the high-order neighbor relationship between samples.To overcome the above challenges,this paper proposes a novel multi-order neighborhood fusion based multi-view deep subspace clustering model.We creatively integrate the multi-order proximity graph structures of different views into the self-expressive layer by a multi-order neighborhood fusion module.By this design,the multi-order Laplacian matrix supervises the learning of the view-consistent self-representation affinity matrix;then,we can obtain an optimal global affinity matrix where each connected node belongs to one cluster.In addition,the discriminative constraint between views is designed to further improve the clustering performance.A range of experiments on six public datasets demonstrates that the method performs better than other advanced multi-view clustering methods.The code is available at https://github.com/songzuolong/MNF-MDSC(accessed on 25 December 2024).
基金funding support from the National Natural Science Foundation of China(Grant No.42007269)the Young Talent Fund of Xi'an Association for Science and Technology(Grant No.959202313094)the Fundamental Research Funds for the Central Universities,CHD(Grant No.300102263401).
文摘The characterization and clustering of rock discontinuity sets are a crucial and challenging task in rock mechanics and geotechnical engineering.Over the past few decades,the clustering of discontinuity sets has undergone rapid and remarkable development.However,there is no relevant literature summarizing these achievements,and this paper attempts to elaborate on the current status and prospects in this field.Specifically,this review aims to discuss the development process of clustering methods for discontinuity sets and the state-of-the-art relevant algorithms.First,we introduce the importance of discontinuity clustering analysis and follow the comprehensive characterization approaches of discontinuity data.A bibliometric analysis is subsequently conducted to clarify the current status and development characteristics of the clustering of discontinuity sets.The methods for the clustering analysis of rock discontinuities are reviewed in terms of single-and multi-parameter clustering methods.Single-parameter methods can be classified into empirical judgment methods,dynamic clustering methods,relative static clustering methods,and static clustering methods,reflecting the continuous optimization and improvement of clustering algorithms.Moreover,this paper compares the current mainstream of single-parameter clustering methods with multi-parameter clustering methods.It is emphasized that the current single-parameter clustering methods have reached their performance limits,with little room for improvement,and that there is a need to extend the study of multi-parameter clustering methods.Finally,several suggestions are offered for future research on the clustering of discontinuity sets.
基金funded by the Catedra Sanitas de Salud y Medio Ambiente of Universidad de Navarrafunded by the Agencia Estatal de Investigatión under the project“Gemelo Digital de Nueva Generatión de Edificios Inteligentes”(DigiTwin)(ref.CPP2021-008909).
文摘European directives advocate for end-users to be aware of their energy consumption.However,individual energy monitoring tools,such as energy meters or cost allocators,are not always affordable or technically feasible to install.Therefore,the development of virtual tools that enable the study of energy consumption in existing buildings is necessary.Virtual sensors,particularly based on white-box models,offer the opportunity to recreate these behaviours.When calibrated with measured data,white-box models,which incorporate detailed building physics,become increasingly valuable for designing energy-efficient buildings.This research explores a novel approach to identifying building’s load period directly from energy data generated by these calibrated models.The volume of data generated by white-box models can be overwhelming for visual analysis,but the hypothesis here is that analysing this data through clustering techniques can reveal patterns related to occupant behaviour and operational schedules.By feeding indoor temperature data into the calibrated model and analysing the resulting energy outputs,the research proposes a method to identify the heating,ventilation and air conditioning(HVAC)system operation schedule,free oscillation periods and non-recurrent events.Validation is achieved by comparing the identified periods with actual measured data.This methodology enables the development of a virtual sensor for cost allocation,which minimises the need for physical sensor deployment while complying with European Union directives.The research not only demonstrates high accuracy but also the potential to outperform measured schedule.This suggests the ability of the method to identify missing sensor data or other factors affecting temperature curves,enabling fault detection and diagnostics(FDD).Consequently,this opens doors for setting optimised operation schedules that balance energy efficiency with occupant comfort.
文摘Reliable Cluster Head(CH)selectionbased routing protocols are necessary for increasing the packet transmission efficiency with optimal path discovery that never introduces degradation over the transmission reliability.In this paper,Hybrid Golden Jackal,and Improved Whale Optimization Algorithm(HGJIWOA)is proposed as an effective and optimal routing protocol that guarantees efficient routing of data packets in the established between the CHs and the movable sink.This HGJIWOA included the phases of Dynamic Lens-Imaging Learning Strategy and Novel Update Rules for determining the reliable route essential for data packets broadcasting attained through fitness measure estimation-based CH selection.The process of CH selection achieved using Golden Jackal Optimization Algorithm(GJOA)completely depends on the factors of maintainability,consistency,trust,delay,and energy.The adopted GJOA algorithm play a dominant role in determining the optimal path of routing depending on the parameter of reduced delay and minimal distance.It further utilized Improved Whale Optimisation Algorithm(IWOA)for forwarding the data from chosen CHs to the BS via optimized route depending on the parameters of energy and distance.It also included a reliable route maintenance process that aids in deciding the selected route through which data need to be transmitted or re-routed.The simulation outcomes of the proposed HGJIWOA mechanism with different sensor nodes confirmed an improved mean throughput of 18.21%,sustained residual energy of 19.64%with minimized end-to-end delay of 21.82%,better than the competitive CH selection approaches.
基金funded by the National Nature Sciences Foundation of China with Grant No.42250410321。
文摘Music recommendation systems are essential due to the vast amount of music available on streaming platforms,which can overwhelm users trying to find new tracks that match their preferences.These systems analyze users’emotional responses,listening habits,and personal preferences to provide personalized suggestions.A significant challenge they face is the“cold start”problem,where new users have no past interactions to guide recommendations.To improve user experience,these systems aimto effectively recommendmusic even to such users by considering their listening behavior and music popularity.This paper introduces a novel music recommendation system that combines order clustering and a convolutional neural network,utilizing user comments and rankings as input.Initially,the system organizes users into clusters based on semantic similarity,followed by the utilization of their rating similarities as input for the convolutional neural network.This network then predicts ratings for unreviewed music by users.Additionally,the system analyses user music listening behaviour and music popularity.Music popularity can help to address cold start users as well.Finally,the proposed method recommends unreviewed music based on predicted high rankings and popularity,taking into account each user’s music listening habits.The proposed method combines predicted high rankings and popularity by first selecting popular unreviewedmusic that themodel predicts to have the highest ratings for each user.Among these,the most popular tracks are prioritized,defined by metrics such as frequency of listening across users.The number of recommended tracks is aligned with each user’s typical listening rate.The experimental findings demonstrate that the new method outperformed other classification techniques and prior recommendation systems,yielding a mean absolute error(MAE)rate and rootmean square error(RMSE)rate of approximately 0.0017,a hit rate of 82.45%,an average normalized discounted cumulative gain(nDCG)of 82.3%,and a prediction accuracy of new ratings at 99.388%.
基金partially supported by the National Natural Science Foundation of China(62161016)the Key Research and Development Project of Lanzhou Jiaotong University(ZDYF2304)+1 种基金the Beijing Engineering Research Center of Highvelocity Railway Broadband Mobile Communications(BHRC-2022-1)Beijing Jiaotong University。
文摘In order to solve the problems of short network lifetime and high data transmission delay in data gathering for wireless sensor network(WSN)caused by uneven energy consumption among nodes,a hybrid energy efficient clustering routing base on firefly and pigeon-inspired algorithm(FF-PIA)is proposed to optimise the data transmission path.After having obtained the optimal number of cluster head node(CH),its result might be taken as the basis of producing the initial population of FF-PIA algorithm.The L′evy flight mechanism and adaptive inertia weighting are employed in the algorithm iteration to balance the contradiction between the global search and the local search.Moreover,a Gaussian perturbation strategy is applied to update the optimal solution,ensuring the algorithm can jump out of the local optimal solution.And,in the WSN data gathering,a onedimensional signal reconstruction algorithm model is developed by dilated convolution and residual neural networks(DCRNN).We conducted experiments on the National Oceanic and Atmospheric Administration(NOAA)dataset.It shows that the DCRNN modeldriven data reconstruction algorithm improves the reconstruction accuracy as well as the reconstruction time performance.FF-PIA and DCRNN clustering routing co-simulation reveals that the proposed algorithm can effectively improve the performance in extending the network lifetime and reducing data transmission delay.
基金Natural Science Foundation of Shandong Province,Grant/Award Number:ZR202103010903Doctoral Fund of Shandong Jianzhu University,Grant/Award Number:X21101Z。
文摘To guarantee safe and efficient tunneling of a tunnel boring machine(TBM),rapid and accurate judgment of the rock mass condition is essential.Based on fuzzy C-means clustering,this paper proposes a grouped machine learning method for predicting rock mass parameters.An elaborate data set on field rock mass is collected,which also matches field TBM tunneling.Meanwhile,target stratum samples are divided into several clusters by fuzzy C-means clustering,and multiple submodels are trained by samples in different clusters with the input of pretreated TBM tunneling data and the output of rock mass parameter data.Each testing sample or newly encountered tunneling condition can be predicted by multiple submodels with the weight of the membership degree of the sample to each cluster.The proposed method has been realized by 100 training samples and verified by 30 testing samples collected from the C1 part of the Pearl Delta water resources allocation project.The average percentage error of uniaxial compressive strength and joint frequency(Jf)of the 30 testing samples predicted by the pure back propagation(BP)neural network is 13.62%and 12.38%,while that predicted by the BP neural network combined with fuzzy C-means is 7.66%and6.40%,respectively.In addition,by combining fuzzy C-means clustering,the prediction accuracies of support vector regression and random forest are also improved to different degrees,which demonstrates that fuzzy C-means clustering is helpful for improving the prediction accuracy of machine learning and thus has good applicability.Accordingly,the proposed method is valuable for predicting rock mass parameters during TBM tunneling.
基金This work is supported by the National Key R&D Program of China(2017YFB0802900).
文摘In recent years,many unknown protocols are constantly emerging,and they bring severe challenges to network security and network management.Existing unknown protocol recognition methods suffer from weak feature extraction ability,and they cannot mine the discriminating features of the protocol data thoroughly.To address the issue,we propose an unknown application layer protocol recognition method based on deep clustering.Deep clustering which consists of the deep neural network and the clustering algorithm can automatically extract the features of the input and cluster the data based on the extracted features.Compared with the traditional clustering methods,deep clustering boasts of higher clustering accuracy.The proposed method utilizes network-in-network(NIN),channel attention,spatial attention and Bidirectional Long Short-term memory(BLSTM)to construct an autoencoder to extract the spatial-temporal features of the protocol data,and utilizes the unsupervised clustering algorithm to recognize the unknown protocols based on the features.The method firstly extracts the application layer protocol data from the network traffic and transforms the data into one-dimensional matrix.Secondly,the autoencoder is pretrained,and the protocol data is compressed into low dimensional latent space by the autoencoder and the initial clustering is performed with K-Means.Finally,the clustering loss is calculated and the classification model is optimized according to the clustering loss.The classification results can be obtained when the classification model is optimal.Compared with the existing unknown protocol recognition methods,the proposed method utilizes deep clustering to cluster the unknown protocols,and it can mine the key features of the protocol data and recognize the unknown protocols accurately.Experimental results show that the proposed method can effectively recognize the unknown protocols,and its performance is better than other methods.
基金sponsored by the National Natural Science Foundation of P.R.China(Nos.62102194 and 62102196)Six Talent Peaks Project of Jiangsu Province(No.RJFW-111)Postgraduate Research and Practice Innovation Program of Jiangsu Province(Nos.KYCX23_1087 and KYCX22_1027).
文摘The study delves into the expanding role of network platforms in our daily lives, encompassing various mediums like blogs, forums, online chats, and prominent social media platforms such as Facebook, Twitter, and Instagram. While these platforms offer avenues for self-expression and community support, they concurrently harbor negative impacts, fostering antisocial behaviors like phishing, impersonation, hate speech, cyberbullying, cyberstalking, cyberterrorism, fake news propagation, spamming, and fraud. Notably, individuals also leverage these platforms to connect with authorities and seek aid during disasters. The overarching objective of this research is to address the dual nature of network platforms by proposing innovative methodologies aimed at enhancing their positive aspects and mitigating their negative repercussions. To achieve this, the study introduces a weight learning method grounded in multi-linear attribute ranking. This approach serves to evaluate the significance of attribute combinations across all feature spaces. Additionally, a novel clustering method based on tensors is proposed to elevate the quality of clustering while effectively distinguishing selected features. The methodology incorporates a weighted average similarity matrix and optionally integrates weighted Euclidean distance, contributing to a more nuanced understanding of attribute importance. The analysis of the proposed methods yields significant findings. The weight learning method proves instrumental in discerning the importance of attribute combinations, shedding light on key aspects within feature spaces. Simultaneously, the clustering method based on tensors exhibits improved efficacy in enhancing clustering quality and feature distinction. This not only advances our understanding of attribute importance but also paves the way for more nuanced data analysis methodologies. In conclusion, this research underscores the pivotal role of network platforms in contemporary society, emphasizing their potential for both positive contributions and adverse consequences. The proposed methodologies offer novel approaches to address these dualities, providing a foundation for future research and practical applications. Ultimately, this study contributes to the ongoing discourse on optimizing the utility of network platforms while minimizing their negative impacts.
基金supported in part by NUS startup grantthe National Natural Science Foundation of China (52076037)。
文摘Although many multi-view clustering(MVC) algorithms with acceptable performances have been presented, to the best of our knowledge, nearly all of them need to be fed with the correct number of clusters. In addition, these existing algorithms create only the hard and fuzzy partitions for multi-view objects,which are often located in highly-overlapping areas of multi-view feature space. The adoption of hard and fuzzy partition ignores the ambiguity and uncertainty in the assignment of objects, likely leading to performance degradation. To address these issues, we propose a novel sparse reconstructive multi-view evidential clustering algorithm(SRMVEC). Based on a sparse reconstructive procedure, SRMVEC learns a shared affinity matrix across views, and maps multi-view objects to a 2-dimensional humanreadable chart by calculating 2 newly defined mathematical metrics for each object. From this chart, users can detect the number of clusters and select several objects existing in the dataset as cluster centers. Then, SRMVEC derives a credal partition under the framework of evidence theory, improving the fault tolerance of clustering. Ablation studies show the benefits of adopting the sparse reconstructive procedure and evidence theory. Besides,SRMVEC delivers effectiveness on benchmark datasets by outperforming some state-of-the-art methods.
文摘Multi-view Subspace Clustering (MVSC) emerges as an advanced clustering method, designed to integrate diverse views to uncover a common subspace, enhancing the accuracy and robustness of clustering results. The significance of low-rank prior in MVSC is emphasized, highlighting its role in capturing the global data structure across views for improved performance. However, it faces challenges with outlier sensitivity due to its reliance on the Frobenius norm for error measurement. Addressing this, our paper proposes a Low-Rank Multi-view Subspace Clustering Based on Sparse Regularization (LMVSC- Sparse) approach. Sparse regularization helps in selecting the most relevant features or views for clustering while ignoring irrelevant or noisy ones. This leads to a more efficient and effective representation of the data, improving the clustering accuracy and robustness, especially in the presence of outliers or noisy data. By incorporating sparse regularization, LMVSC-Sparse can effectively handle outlier sensitivity, which is a common challenge in traditional MVSC methods relying solely on low-rank priors. Then Alternating Direction Method of Multipliers (ADMM) algorithm is employed to solve the proposed optimization problems. Our comprehensive experiments demonstrate the efficiency and effectiveness of LMVSC-Sparse, offering a robust alternative to traditional MVSC methods.
基金This work was supported by Science and Technology Research Program of Chongqing Municipal Education Commission(KJZD-M202300502,KJQN201800539).
文摘In clustering algorithms,the selection of neighbors significantly affects the quality of the final clustering results.While various neighbor relationships exist,such as K-nearest neighbors,natural neighbors,and shared neighbors,most neighbor relationships can only handle single structural relationships,and the identification accuracy is low for datasets with multiple structures.In life,people’s first instinct for complex things is to divide them into multiple parts to complete.Partitioning the dataset into more sub-graphs is a good idea approach to identifying complex structures.Taking inspiration from this,we propose a novel neighbor method:Shared Natural Neighbors(SNaN).To demonstrate the superiority of this neighbor method,we propose a shared natural neighbors-based hierarchical clustering algorithm for discovering arbitrary-shaped clusters(HC-SNaN).Our algorithm excels in identifying both spherical clusters and manifold clusters.Tested on synthetic datasets and real-world datasets,HC-SNaN demonstrates significant advantages over existing clustering algorithms,particularly when dealing with datasets containing arbitrary shapes.
基金supported in part by the National Natural Science Foundation of China under Grant 62171203in part by the Jiangsu Province“333 Project”High-Level Talent Cultivation Subsidized Project+2 种基金in part by the SuzhouKey Supporting Subjects for Health Informatics under Grant SZFCXK202147in part by the Changshu Science and Technology Program under Grants CS202015 and CS202246in part by Changshu Key Laboratory of Medical Artificial Intelligence and Big Data under Grants CYZ202301 and CS202314.
文摘In this paper,we introduce a novel Multi-scale and Auto-tuned Semi-supervised Deep Subspace Clustering(MAS-DSC)algorithm,aimed at addressing the challenges of deep subspace clustering in high-dimensional real-world data,particularly in the field of medical imaging.Traditional deep subspace clustering algorithms,which are mostly unsupervised,are limited in their ability to effectively utilize the inherent prior knowledge in medical images.Our MAS-DSC algorithm incorporates a semi-supervised learning framework that uses a small amount of labeled data to guide the clustering process,thereby enhancing the discriminative power of the feature representations.Additionally,the multi-scale feature extraction mechanism is designed to adapt to the complexity of medical imaging data,resulting in more accurate clustering performance.To address the difficulty of hyperparameter selection in deep subspace clustering,this paper employs a Bayesian optimization algorithm for adaptive tuning of hyperparameters related to subspace clustering,prior knowledge constraints,and model loss weights.Extensive experiments on standard clustering datasets,including ORL,Coil20,and Coil100,validate the effectiveness of the MAS-DSC algorithm.The results show that with its multi-scale network structure and Bayesian hyperparameter optimization,MAS-DSC achieves excellent clustering results on these datasets.Furthermore,tests on a brain tumor dataset demonstrate the robustness of the algorithm and its ability to leverage prior knowledge for efficient feature extraction and enhanced clustering performance within a semi-supervised learning framework.
基金supported by the National Key Research and Development Program of China(No.2022YFB3304400)the National Natural Science Foundation of China(Nos.6230311,62303111,62076060,61932007,and 62176083)the Key Research and Development Program of Jiangsu Province of China(No.BE2022157).
文摘Traditional Fuzzy C-Means(FCM)and Possibilistic C-Means(PCM)clustering algorithms are data-driven,and their objective function minimization process is based on the available numeric data.Recently,knowledge hints have been introduced to formknowledge-driven clustering algorithms,which reveal a data structure that considers not only the relationships between data but also the compatibility with knowledge hints.However,these algorithms cannot produce the optimal number of clusters by the clustering algorithm itself;they require the assistance of evaluation indices.Moreover,knowledge hints are usually used as part of the data structure(directly replacing some clustering centers),which severely limits the flexibility of the algorithm and can lead to knowledgemisguidance.To solve this problem,this study designs a newknowledge-driven clustering algorithmcalled the PCM clusteringwith High-density Points(HP-PCM),in which domain knowledge is represented in the form of so-called high-density points.First,a newdatadensitycalculation function is proposed.The Density Knowledge Points Extraction(DKPE)method is established to filter out high-density points from the dataset to form knowledge hints.Then,these hints are incorporated into the PCM objective function so that the clustering algorithm is guided by high-density points to discover the natural data structure.Finally,the initial number of clusters is set to be greater than the true one based on the number of knowledge hints.Then,the HP-PCM algorithm automatically determines the final number of clusters during the clustering process by considering the cluster elimination mechanism.Through experimental studies,including some comparative analyses,the results highlight the effectiveness of the proposed algorithm,such as the increased success rate in clustering,the ability to determine the optimal cluster number,and the faster convergence speed.
文摘Path-based clustering algorithms typically generate clusters by optimizing a benchmark function.Most optimiza-tion methods in clustering algorithms often offer solutions close to the general optimal value.This study achieves the global optimum value for the criterion function in a shorter time using the minimax distance,Maximum Spanning Tree“MST”,and meta-heuristic algorithms,including Genetic Algorithm“GA”and Particle Swarm Optimization“PSO”.The Fast Path-based Clustering“FPC”algorithm proposed in this paper can find cluster centers correctly in most datasets and quickly perform clustering operations.The FPC does this operation using MST,the minimax distance,and a new hybrid meta-heuristic algorithm in a few rounds of algorithm iterations.This algorithm can achieve the global optimal value,and the main clustering process of the algorithm has a computational complexity of O�k2×n�.However,due to the complexity of the minimum distance algorithm,the total computational complexity is O�n2�.Experimental results of FPC on synthetic datasets with arbitrary shapes demonstrate that the algorithm is resistant to noise and outliers and can correctly identify clusters of varying sizes and numbers.In addition,the FPC requires the number of clusters as the only parameter to perform the clustering process.A comparative analysis of FPC and other clustering algorithms in this domain indicates that FPC exhibits superior speed,stability,and performance.
基金supported by the Key Research and Development Program of Hainan Province(Grant Nos.ZDYF2023GXJS163,ZDYF2024GXJS014)National Natural Science Foundation of China(NSFC)(Grant Nos.62162022,62162024)+3 种基金the Major Science and Technology Project of Hainan Province(Grant No.ZDKJ2020012)Hainan Provincial Natural Science Foundation of China(Grant No.620MS021)Youth Foundation Project of Hainan Natural Science Foundation(621QN211)Innovative Research Project for Graduate Students in Hainan Province(Grant Nos.Qhys2023-96,Qhys2023-95).
文摘Contrastive learning is a significant research direction in the field of deep learning.However,existing data augmentation methods often lead to issues such as semantic drift in generated views while the complexity of model pre-training limits further improvement in the performance of existing methods.To address these challenges,we propose the Efficient Clustering Network based on Matrix Factorization(ECN-MF).Specifically,we design a batched low-rank Singular Value Decomposition(SVD)algorithm for data augmentation to eliminate redundant information and uncover major patterns of variation and key information in the data.Additionally,we design a Mutual Information-Enhanced Clustering Module(MI-ECM)to accelerate the training process by leveraging a simple architecture to bring samples from the same cluster closer while pushing samples from other clusters apart.Extensive experiments on six datasets demonstrate that ECN-MF exhibits more effective performance compared to state-of-the-art algorithms.
