Fuzzy clustering algorithm with locality preserving based on anchor graph

Consistency-Regularized Multi-Stage Joint-Perception Graph Fuzzy Clustering Algorithm

Abstract Individual cognitive differences may cause decision makers to interpret the same linguistic terms differently. To address this issue, this paper proposes a novel consensus model for large-scale group decision-making by incorporating personalized individual semantics into a probabilistic linguistic preference framework. A normalization method integrating emotional tone is introduced to refine probabilistic linguistic preference relations, and an additive consistency-based semantic optimization model is developed to assign appropriate linguistic terms to decision makers. To promote interaction among those with similar interests, a weighting method based on semantic similarity and a fuzzy clustering algorithm using personalized individual semantics are employed to form subgroups with similar semantics. A consensus-reaching process, including assessment and feedback stages, is then applied to guide decision makers toward agreement. A case study on environmental project selection verifies the effectiveness and applicability of the proposed approach.

This study proposes Tsallis entropy-regularized Yang-type fuzzy c -hidden Markov models (TYFCHMMs), a fuzzy clustering algorithm based on hidden Markov model (HMM), for series data. First, the relationship between a Gaussian mixture model with identity covariances, which is a conventional probabilistic clustering algorithm for vectorial data, and Yang-type fuzzy c -means, which is a conventional fuzzy clustering algorithm for vectorial data, is determined. Second, the relationship between Bezdek-type fuzzy c -means and Tsallis entropy-regularized fuzzy c -means, which are two conventional fuzzy clustering algorithms for vectorial data, is determined. Based on these relationships, TYFCHMMs are constructed from mixtures of HMMs (MoHMMs), which are conventional probabilistic clustering algorithms based on HMM for series data, using Yang-type fuzzification and Tsallis entropy regularization. Through numerical tests using an artificial dataset, the effects of parameters on the clustering results of TYFCHMMs and the close relationship between TYFCHMMs and MoHMMs are identified. Furthermore, numerical tests using ten real datasets confirmed that TYFCHMMs outperform MoHMMs in terms of clustering accuracy.

Thermal error is an important obstacle to improve the thermal stability and accuracy of machine tools. A theoretical model of heat transfer and conduction for the high-speed electric spindle of CNC machine tools was established by simplifying the electric spindle into a one-dimensional rod for heat transfer, proving the hysteresis thermal characteristics of the electric spindle. To solve difficulty in predicting machine tool spindles under variable operating conditions. A fuzzy C-means clustering (FCM) algorithm and uncertainty coefficient method (UCM) were proposed to determine temperature sensitive points, and the influence of rotational speed on thermal error was considered to establish a nonlinear autoregressive (NAR) long short-term memory (LSTM) neural network model. The use of Sparrow search algorithm (SSA) for automatic optimization of hyperparameters in the NAR-LSTM model improves the prediction accuracy and modeling efficiency of modeling. Experiments results shows thermal error of the spindle is arisen from 0 to 65 μ m under complex machining conditions, the comprehensive predictive residual of the SSA-NAR-LSTM, LSTM and BP models under variable conditions remained 3, 6.8 and 8 μ m . The predictive accuracy of the SSA-NAR-LSTM model increased over 50% compared to BP and LSTM models. The proposed SSA-NAR-LSTM modeling method is a high-precision and effective method for predicting thermal errors in high-speed electric spindles.

Short-term peak shaving model for a wind-solar-pumped hydropower storage system fully using storage flexibility by dynamic fuzzy clustering algorithm

ABSTRACT Train speed control is crucial for automatic train operation (ATO), but challenges remain in modeling and control due to the nonlinear, time-varying, and coupled nature of train operation. This paper proposes an improved T-S model-based train speed generalized predictive control (GPC) method. Compared with previous works, this approach establishes a more accurate dynamic model using complex train operation data, providing a more precise parameter prediction model for GPC to enhance control performance. First, a generalized universal adaptive fuzzy c-means (GUA_FCM) algorithm is proposed, overcoming traditional fuzzy clustering algorithm limitations by constructing a novel objective function. A GUA_FCM-based improved Takagi-Sugeno (T-S) model is then established, enhancing the train model identification accuracy and providing a better parameter prediction model for GPC. Then, an improved T-S model-based train speed GPC method is proposed. This method predicts future system behavior, optimizes control decisions, and improves speed tracking accuracy. Finally, experimental results confirm its effectiveness.

Digital courses supported by traditional resource construction and sharing methods are difficult to meet the needs of learners at all levels and provide continuous long-term learning support services. Therefore, this paper puts forward a collaborative filtering (CF) recommendation method based on user influence relationship and constructs a personalized recommendation model for college students' educational resources so as to realize the all-round education and growth of college students. The research results show that the personalized recommendation model has low error and high recommendation accuracy. Compared with fuzzy clustering algorithm (FCA), the error is reduced by 18.88% and the recall rate is increased by 23.86%. Educators should pay attention to the control and screening of personalized recommendation content on new media platforms; ensure that positive, objective, and high-quality information resources are delivered to college students; and help them establish correct values.

Abstract This study proposes an integrated analytical framework to enhance football performance analytics by combining feature engineering, fuzzy clustering, interpretable machine learning, and topological network analysis. The framework is designed to extract latent offensive profiles and predict high-efficiency scoring profiles across domestic and international competitions. The approach begins by constructing three composite indicators – Index of Offensive Efficiency, Competitive Resilience Index, and Versatility Score – designed to capture multidimensional aspects of a player’s offensive productivity, adaptability across competitions, and contribution breadth. These engineered metrics inform a fuzzy clustering algorithm that reveals two core performance profiles: “Seasoned Finishing Specialists” and “Emerging Versatile Contributors”. Building on this segmentation, a supervised learning model based on XGBoost is employed to predict the likelihood of surpassing a goals-per-shot efficiency threshold. Model interpretability is ensured via SHAP plot, which highlight the pivotal role of salary, finishing metrics, and competition-specific resilience. Partial dependence plots further expose nonlinear and interactive effects between key predictors. A network-based analysis complements the model by mapping performance similarities and identifying both archetypal and transitional performers via centrality measures. Robustness checks, including alternative winsorization, fuzziness levels, and subgroup-specific clustering, confirm the stability of the results. Overall, the proposed framework bridges segmentation and prediction with transparency and domain-relevance, offering a comprehensive toolkit for decision-makers in sports analytics, recruiters, and talent management.

Abstract The fuzzy c-means (FCM) algorithm is widely used image segmentation but, has several limitations. It is sensitive to noise, demonstrates variable convergence rate depending on data distribution, and its reliance on Euclidean distance fails to account for intra-cluster variations, particularly in complex and color images. Furthermore, FCM’s non-adaptive distance metric struggles with diverse cluster shapes, and most FCM-based approaches face difficulties in color image segmentation due to the challenges in spatial information acquisition. To address these limitations, we propose an Improved Gustafson-Kessel (IGK) algorithm that offers superior robustness compared to both FCM and traditional Gustafson-Kessel (GK) clustering. Our approach first applies morphological reconstruction (MR) for grayscale images and multivariate morphological reconstruction (MMR) for color images to ensure noise immunity while preservation image details. We then replace the Euclidean distance metric with Mahalanobis distance to adapt to varying cluster shapes. The algorithm iteratively updates cluster centers, membership matrix, and positive definite symmetric matrices, followed by a median filter refinement of the membership partition matrix. Unlike previous approaches, IGK eliminates the need for computing distances within local spatial neighbors during clustering. Experimental results on both grayscale and color images demonstrate that the proposed IGK algorithm achieves superior segmentation performance compared to existing FCM-based methods.

Abstract The optimization of the logistics distribution path is an important link to improve the distribution efficiency, reduce the cost, and realize the green logistics. However, the logistics distribution scenario is complex and changeable, facing multiple challenges of multi-objective optimization, multi-constraints and dynamic environment. To solve the above problems, this paper proposes an optimization model of logistics distribution path based on fuzzy clustering and multi-objective optimization. By collecting and analyzing the actual distribution data, the model uses fuzzy clustering algorithm to divide the distribution points, and decomposes the global optimization problem into multiple local problems, thus reducing the computational complexity. Subsequently, the multi-target optimization algorithm was used to balance the path length, delivery time and carbon emissions targets, significantly improving the distribution efficiency. The results show that the optimization model proposed here shows significant advantages in path length, distribution time, and carbon emission, and also has a significantly shorter computation time. In addition, this paper also discusses the practical application scenarios of the optimization model, including urban logistics, emergency logistics and cross-border e-commerce logistics, etc Future research can further enhance the dynamic adaptability of the model, introduce real-time optimization technology, and combine the Internet of Things and unmanned distribution technology to provide stronger technical support for the development of intelligent logistics.

Abstract Research on work accidents is important to determine the causes of occupational accidents to effectively prevent them in the future and improve workplace safety. This study aims to analyse the evolution of accidents in the metal products manufacturing subsector for construction (CNAE subsector 251) in Spain for the period 2009–2022 to classify accidents into different operational profiles. This will facilitate the proposal of specific preventive measures based on the severity and characteristics of each accident. Data for this study are collected from occupational accident reports via the Delt@ (Electronic declaration of injured workers) IT system. The study variables were classified into five groups: temporal, personal, business, circumstances, and consequences. Accidents at work are more common in males and in middle-aged workers (30–59 years). Companies with less than five workers, works outside the usual workplace and workers with less three months of length of the service in the company present high accident rate, both in light as fatal accidents. A semi-supervised model has been developed using the Fuzzy Cluster algorithm that can detect serious accidents with a recall rate of approximately 64% and group them into three distinct categories. This makes it possible to propose specific preventive measures for each category, of which there are 12 in total.

Abstract This paper addresses the ambitious goal of merging two different approaches to group detection in complex domains: one based on fuzzy clustering and the other on community detection theory. To achieve this, two clustering algorithms are proposed: fuzzy C-medoids clustering with modularity spatial correction and fuzzy C-modes clustering with modularity spatial correction. The former is designed for quantitative data, while the latter is intended for qualitative data. The concept of fuzzy modularity is introduced into the standard objective function of fuzzy clustering algorithms as a spatial regularization term, whose contribution to the clustering criterion based on attributes is controlled by an exogenous parameter. An extensive simulation study is conducted to support the theoretical framework, complemented by two applications to real-world data related to the theme of sustainability. The first application involves data from the 2030 Agenda for Sustainable Development, while the second focuses on urban green spaces in Italian provincial capitals and metropolitan cities. Both the simulation results and the applications demonstrate the advantages of this new methodological proposal.

The large-scale integration of coordinated offshore wind and offshore photovoltaic (PV) generation introduces pronounced power fluctuations due to the intrinsic randomness and intermittency of renewable energy sources (RESs). These fluctuations pose significant challenges to the secure, stable, and economical operation of modern power systems. To address this issue, this study proposes a hybrid energy storage system (HESS)-based optimization framework that simultaneously enhances fluctuation suppression performance, optimizes storage capacity allocation, and improves life-cycle economic efficiency. First, a K-means fuzzy clustering algorithm is employed to analyze historical RES power data, extracting representative daily fluctuation profiles to serve as accurate inputs for optimization. Second, the time-varying filter empirical mode decomposition (TVF-EMD) technique is applied to adaptively decompose the net power fluctuations. High-frequency components are allocated to a flywheel energy storage system (FESS), valued for its high power density, rapid response, and long cycle life, while low-frequency components are assigned to a battery energy storage system (BESS), characterized by high energy density and cost-effectiveness. This decomposition–allocation strategy fully exploits the complementary characteristics of different storage technologies. Simulation results for an integrated offshore wind–PV generation scenario demonstrate that the proposed method significantly reduces the fluctuation rate of RES power output while maintaining favorable economic performance. The approach achieves unified optimization of HESS sizing, fluctuation mitigation, and life-cycle cost, offering a viable reference for the planning and operation of large-scale offshore hybrid renewable plants.

Most clustering validity indexes (CVIs) for fuzzy clustering are based upon the fuzzy c-means (FCMs) algorithm, and the effect of these CVIs is limited due to the "uniform effect" of FCM. Besides, main existing CVIs have the problems of incompleteness characterization of separateness and weak performance for noisy datasets. To address these challenges, the multi-granularity fusion (MGF) index is proposed. First, MGF synthetically considers the FCM, possibilistic fuzzy c-means and kernel-based FCM algorithms, which is more comprehensive than just considering FCM. Second, we add a perturbation to the sum of the partition matrix as the fuzzy cardinality and combine it with the fuzzy weighted distance, which are helpful to grasp the compactness. Third, four elements are considered together to characterize the separateness, incorporating the minimum distance, the maximum distance, the mean distance, and the sample variance of cluster center, where the last one can make the separateness unbiased from the macroscopic perspective. Besides, the convergence of MGF is proved. Finally, we test MGF for five algorithms on 36 datasets comparing with 14 CVIs, validating the accuracy and stability of MGF. It is observed that MGF can get superior results than other CVIs, especially for high-dimensional datasets and noisy datasets.

ABSTRACT Accurate precipitation prediction is vital for effective water resource management, agricultural planning, and natural disaster mitigation. Traditional forecasting methods often encounter difficulties due to the nonlinearity, complex seasonality, and noise inherent in meteorological data. This paper introduces a novel methodology that combines the FB–Prophet algorithm, designed by Facebook for identifying trends and seasonal patterns, with a fuzzy clustering algorithm. This integration aims to refine a crucial aspect of the FB–Prophet framework: the identification and incorporation of special events, specifically holidays, which play a significant role in the predictive modeling process. This approach ensures that holidays are effectively integrated into forecasts, enhancing the model's overall accuracy and reliability. Additionally, the proposed model is compared to several widely used algorithms in recent studies in terms of accuracy, employing nonparametric tests for a robust evaluation. Empirical results demonstrate a significant improvement in forecast accuracy over traditional methods.

Customer segmentation allows companies to create mutual profiles of their customers. Determining industrial customer segments based on a single perspective causes various customer features to be disregarded. This study aims to develop a holistic segmentation approach in a B2B setting. The paper proposes a multi-dimensional segmentation model with four main criteria: customer purchasing performance, customer cooperation, customer workload, and customer potential. The case study demonstrates the real-life application of the proposed model using 379 customer data and 17 sub-criteria under four dimensions. The Fuzzy C-Means Clustering Algorithm creates the customer segments, and the Fuzzy Analytical Hierarchical Process is used to calculate criteria weights. The marketing strategies of each segment are used to guide customer relations and managerial decisions. This paper suggests that companies segment their customers by considering financial performance, cooperation level, future potential throughput, and challenges. It provides a practical and holistic insight into industrial customer segmentation.

Abstract Understanding customer behavior has become critical for marketers and tourist service providers since global travel has rapidly expanded. Recognizing consumer behavior patterns allows organizations to modify their strategy for better service delivery. However, the conventional approaches frequently fail to adequately capture the variety and complexity of tourism consumer behavior because of the large and diverse data available. This research seeks to investigate the use of fuzzy clustering analysis to better understand tourism consumer behavior patterns. The method combines fuzzy clustering algorithms with customer behavior data such as demographics, travel preferences, and purchasing patterns. The investigation reveals separate groups of consumers, providing insights into how various factors influence tourist purchasing decisions. The data were gathered using questionnaires, online booking platforms, and travel websites, where customers provided information about their previous travel experiences and preferences. Data preparation was used to normalize the data for analysis. Principal component analysis was employed to decrease dimensionality. The Sea Turtle Foraging Optimized Fuzzy C-Means clustering (STFO-FCMC) is presented as an extension of normal FCMC that incorporates an optimization procedure based on sea turtle foraging habits. This optimization enhances the accuracy and efficiency of cluster center selection and membership values, making STFO-FCMC especially well-suited for dealing with the complexity and unpredictability of tourism behavior data. The findings show multiple consumer behavior patterns, including diverse preferences for various types of tourist products and services, which are split by age, income, and travel objectives. The STFO-FCMC method is assessed using metrics, including accuracy of 97.84%, precision, recall, and F1-score. These data assist service providers create individualized services and marketing strategies that improve consumer satisfaction and business performance. Overall, fuzzy clustering analysis, particularly with the STFO-FCMC approach, is a successful tool for detecting tourist consumer behavior, with substantial promise for improving tourism product and service targeting.

Traditional fuzzy clustering algorithms construct sample partition criteria solely based on similarity measures but lack an effective representation of prior membership information, which limits further improvements in clustering accuracy. To address this issue, this paper proposes a semi-supervised fuzzy clustering algorithm based on prior membership (SFCM-PM). The proposed algorithm introduces prior information entropy as a metric to quantify the divergence between partition membership and prior membership and incorporates this as an auxiliary partition criterion into the objective function. By jointly optimizing data similarity and consistency with prior knowledge during the clustering process, the algorithm achieves more accurate and reliable clustering results. The experimental results demonstrate that the SFCM-PM algorithm achieves significant performance improvements by incorporating a small number of prior membership samples across several standard and real-world datasets. It also performs outstandingly on datasets with unbalanced sample distributions.

Focusing on the issue of loss of image spatial structure information in the current oil painting style recognition research, this paper first preprocesses the oil painting image, then combines the spatial information and grayscale information to improve the fuzzy clustering algorithm (FCM) (EFCM), and utilizes the EFCM to segment the oil painting image. On this basis, the fourth-order tensor training samples are constructed for the segmented oil painting images, and the oil painting image features are extracted using multilinear principal component analysis (MPCA). Finally, the support vector machine (SVM) optimized by the improved particle swarm algorithm (EAPSO) is used for oil painting style recognition. The simulation outcome demonstrates that the offered model has an average recognition accuracy (mAP) of 96.2%, which is better than the comparison model, and provides a new technical path for the accurate recognition of oil painting style.