Network Clustering Research Articles

Structured Deep Graph Clustering Network Based on Consistency Constraint

Structured Deep Graph Clustering Network Based on Consistency Constraint

DyCR: A Dynamic Clustering and Recovering Network for Few-Shot Class-Incremental Learning.

Few-shot class-incremental learning (FSCIL) aims to continually learn novel data with limited samples. One of the major challenges is the catastrophic forgetting problem of old knowledge while training the model on new data. To alleviate this problem, recent state-of-the-art methods adopt a well-trained static network with fixed parameters at incremental learning stages to maintain old knowledge. These methods suffer from the poor adaptation of the old model with new knowledge. In this work, a dynamic clustering and recovering network (DyCR) is proposed to tackle the adaptation problem and effectively mitigate the forgetting phenomena on FSCIL tasks. Unlike static FSCIL methods, the proposed DyCR network is dynamic and trainable during the incremental learning stages, which makes the network capable of learning new features and better adapting to novel data. To address the forgetting problem and improve the model performance, a novel orthogonal decomposition mechanism is developed to split the feature embeddings into context and category information. The context part is preserved and utilized to recover old class features in future incremental learning stages, which can mitigate the forgetting problem with a much smaller size of data than saving the raw exemplars. The category part is used to optimize the feature embedding space by moving different classes of samples far apart and squeezing the sample distances within the same classes during the training stage. Experiments show that the DyCR network outperforms existing methods on four benchmark datasets. The code is available at: https://github.com/zichengpan/DyCR.

Topological state-space estimation of functional human brain networks.

We introduce an innovative, data-driven topological data analysis (TDA) technique for estimating the state spaces of dynamically changing functional human brain networks at rest. Our method utilizes the Wasserstein distance to measure topological differences, enabling the clustering of brain networks into distinct topological states. This technique outperforms the commonly used k-means clustering in identifying brain network state spaces by effectively incorporating the temporal dynamics of the data without the need for explicit model specification. We further investigate the genetic underpinnings of these topological features using a twin study design, examining the heritability of such state changes. Our findings suggest that the topology of brain networks, particularly in their dynamic state changes, may hold significant hidden genetic information.

Stripe-schedule Aware Repair in Erasure-coded Clusters with Heterogeneous Star Networks

More and more storage systems use erasure code to tolerate faults. It takes pieces of data blocks as input and encodes a small number of parity blocks as output, where these blocks form a stripe. When reconsidering the recovery problem in the multi-stripe level and heterogeneous network clusters, quickly generating an efficient multi-stripe recovery solution that reduces recovery time remains a challenging and time-consuming task. Previous works either use a greedy algorithm that may fall into the local optimal and have low recovery performance or a meta-heuristic algorithm with a long running time and low solution generation efficiency. In this paper, we propose a Stripe-schedule Aware Repair (SARepair) technique for multi-stripe recovery in heterogeneous erasure-coded clusters based on RS code. By carefully examining the metadata of blocks, SARepair intelligently adjusts the recovery solution for each stripe and obtains another multi-stripe solution with less recovery time in a computationally efficient manner. It then tolerates worse solutions to overcome the local optimal and uses a rollback mechanism to adjust search regions to reduce recovery time further. Moreover, instead of reading blocks sequentially from each node, SARepair also selectively schedules the reading order for each block to reduce the memory overhead. We extend SARepair to address the full-node recovery and adapt to the LRC code. We prototype SARepair and show via both simulations and Amazon EC2 experiments that the recovery performance can be improved by up to 59.97% over a state-of-the-art recovery approach while keeping running time and memory overhead low.

Collaborative Video Caching in the Edge Network using Deep Reinforcement Learning

With the enormous growth in mobile data traffic over the 5G environment, Adaptive BitRate (ABR) video streaming has become a challenging problem. Recent advances in Mobile Edge Computing (MEC) technology make it feasible to use Base Stations (BSs) intelligently by network caching, popularity-based video streaming, etc. Additional computing resources on the edge node offer an opportunity to reduce network traffic on the backhaul links during peak traffic hours. More recently, it has been found in the literature that collaborative caching strategies between neighbouring BSs (i.e., MEC servers) make it more efficient to reduce backhaul traffic and network congestion and thus improve the viewer experience substantially. In this work, we propose a Reinforcement Learning (RL) based collaborative caching mechanism where the edge servers cooperate to serve the requested content from the end-users. Specifically, this research aims to improve the overall cache hit rate at the MEC, where the edge servers are clustered based on their geographic locations. The said task is modelled as a multi-objective optimization problem and solved using an RL framework. In addition, a novel cache admission and eviction policy is defined by calculating the priority score of video segments in the clustered MEC mesh network.

Strategic Planning of Rural Telecommunication Infrastructure: A Multi-Source Data Fusion and Optimization Model

Abstract. The United Nations has underscored the critical role of digital connectivity by integrating it into several sustainable development goals, with the ambition for nations worldwide to achieve comprehensive access by 2030. Over thousand million people, primarily in rural areas, are disconnected from the digital world, highlighting the urgent need for viable and sustainable telecommunications solutions. These areas are characterized by sparse populations and lower economic levels, presenting great challenges for connectivity. This work introduces a strategy for enhancing rural telecommunication planning using geospatial and remote sensing data, deep learning-based clustering techniques, network graphs, and terrain analysis. The objective is to develop an optimal network topology and identify prime locations for telecommunications infrastructure, such as towers or relay stations. The methodology begins with the application of an adapted Deep Embedded Clustering (DEC) technique to identify community boundaries accurately. Then, it combines geospatial data (such as roads, terrain slope, flatness, etc.) and remote sensing data (vegetation, waterways, etc.) through an optimization algorithm. This process aims to determine the most suitable sites for infrastructure placement and the best network topology for connecting these areas. The study focuses on the region of Congo, offering a detailed case study on the application of this approach. Experimental results are presented to demonstrate the effectiveness of the proposed telecommunications expansion strategy.

Software effort estimation using convolutional neural network and fuzzy clustering

Software effort estimation using convolutional neural network and fuzzy clustering

An optimization of IoT usage real-time traffic in smart network clustering

An optimization of IoT usage real-time traffic in smart network clustering

Salt stimulates sulfide−driven autotrophic denitrification: Microbial network and metagenomics analyses

Sulfur autotrophic denitrification (SADN) is a promising biological wastewater treatment technology for nitrogen removal, and its performance highly relies on the collective activities of the microbial community. However, the effect of salt (a prevailing characteristic of some nitrogen−containing industrial wastewaters) on the microbial community of SADN is still unclear. In this study, the response of the sulfide−SADN process to different salinities (i.e., 1.5 % salinity, 0.5 % salinity, and without salinity) as well as the involved microbial mechanisms were investigated by molecular ecological network and metagenomics analyses. Results showed that the satisfactory nitrogen removal efficiency (>97 %) was achieved in the sulfide−SADN process (S/N molar ratio of 0.88) with 1.5 % salinity. In salinity scenarios, the genus Thiobacillus significantly proliferated and was detected as the dominant sulfur−oxidizing bacteria in the sulfide−SADN system, occupying a relative abundance of 29.4 %. Network analysis further elucidated that 1.5 % salinity had enabled the microbial community to form a more densely clustered network, which intensified the interactions between microorganisms and effectively improved the nitrogen removal performance of the sulfide−SADN. Metagenomics sequencing revealed that the abundance of functional genes encoding for key enzymes involved in SADN, dissimilatory nitrate reduction to ammonium, and nitrification was up−regulated in the 1.5 % salinity scenario compared to that without salinity, stimulating the occurrence of multiple nitrogen transformation pathways. These multi−paths contributed to a robust SADN process (i.e., nitrogen removal efficiency >97 %, effluent nitrogen <2.5 mg N/L). This study deepens our understanding of the effect of salt on the SADN system at the community and functional level, and favors to advance the application of this sustainable bioprocess in saline wastewater treatment.

Knowledge-based system and expectation-maximization to discovering causes of imperfect labels in vehicular networks clustering

This paper focuses on the development of a knowledge-based system for automatically diagnosing issues in Vehicular ad hoc networks (VANETs). VANETs enable communication between vehicles and infrastructure, enhancing road safety and efficiency through timely information exchange. The proposed system aims to efficiently maintain and ensure the continuity of network service by leveraging innovative pattern recognition methods tailored to VANETs. The automatic diagnosis problem in VANETs involves estimating the operating class of network components based on sensor observations. This entails associating sensor measurements with specific operating modes. By implementing condition-based preventive maintenance procedures, potential component failures can be detected early, mitigating network disruptions. Various approaches, such as expert systems, fault trees, network state models, and statistical learning through pattern recognition, can be employed to address this problem. This paper primarily focuses on the statistical learning approach, where a classification or regression function is learned from a set of examples to assign operation modes to new measurements. It discusses relevant metrics and preprocessing techniques to simplify the decision-making process. The diagnostic system's results are determined based on the formulation of the classification or regression problem. The learning base is constructed, and an appropriate classification method is selected to develop and validate the automatic diagnosis system. While non-parametric models like support vector machines are commonly used, this article emphasizes the significance of considering assumptions and leveraging additional information to enhance performance. It proposes a more specific formalization of the problem, integrating the unique characteristics of VANETs. The contributions of this article revolve around the theory of belief functions, a generative approach, and the utilization of parametric models defined using graphical models. Experimental studies conducted on artificial datasets have demonstrated the benefits of the semi-supervised approach within the context of VANET networks.

Updated research trend and clustering algorithm on virtual reality and pulmonary rehabilitation: Scopus-based bibliometric and visual analysis

Background: Virtual reality (VR) is a new innovative technology that can enhance intervention and should promote the effectiveness of rehabilitation, but there is a lack of scientific evidence on the clustering and topic research trend, especially on VR and pulmonary rehabilitation (PR). More evidence about network clusters and trends will encourage the research in the future. Objective: This study aimed to explore, identify, cluster, and forecast analysis for the research trend of VR and PR from the research articles on the SCOPUS database. Materials and methods: In this study, the search terms “virtual reality” AND “pulmonary rehabilitation” were extracted from specific English research articles published on the SCOPUS database between 2013-2023. RStudio software was used to perform bibliometric and visual analysis. During the analysis in the bibliometric tool, the normalization process with Salton’s Cosine and network clustering via trend topic with the Walktrap algorithm was analyzed before specific visualization with the network clustering mapping, Treemap, and trend topic line by KamadaKawai layout algorithm. Results: From the 1,396 articles on “VR” published between 2010 and 2023, there were 13 research articles on “VR AND PR” published between 2013 and 2023. The bibliometric result from 13 articles showed total of 36 subdisciplines correlated networks among virtual reality (20, 7%), male (16, 6%), female (15, 5%), aged (12, 4%), chronic obstructive lung disease (12, 4%), exercise (11, 4%), human (10, 4%), humans (10, 4%), middle-aged (10, 4%), quality of life (10, 4%), article (9, 3%), pulmonary rehabilitation (9, 3%), controlled study (8, 3%), adult (6, 2%), chronic obstructive (6, 2%), clinical article (6, 2%), forced expiratory volume (6, 2%), pulmonary disease (6, 2%), randomized controlled trial (6, 2%), and forced vital capacity (5, 2%), etc., respectively. Three network clusters were reported after the normalization process and clustering evaluation by factorial analysis. The first cluster was composed of virtual reality, male, female, aged, chronic obstructive lung disease, exercise, human, humans, middle-aged, quality of life, article, pulmonary rehabilitation, controlled study, adult, chronic obstructive, clinical article, forced expiratory volume, randomized controlled trial, forced vital capacity, six-minute walk test, depression, technology, telerehabilitation, breathing exercise, Covid-19, dyspnea, functional status, hospital patient, and physical activity, respectively. The second cluster consisted of procedure and exercise therapy, and the last cluster consisted of exercise tolerance, lung, treatment outcome, health program, and convalescence. Finally, trend research topics were presented in virtual reality, male, female, aged, chronic obstructive lung disease, human, exercise, quality of life, and middle-aged, respectively, in 2023. Conclusion: Therefore, the contribution from data analysis in this article can identify the clustering and trend topics of VR, chronic obstructive lung disease, aging participants, exercise, and quality of life in future research.

Composite attention mechanism network for deep contrastive multi-view clustering

Contrastive learning-based deep multi-view clustering methods have become a mainstream solution for unlabeled multi-view data. These methods usually utilize a basic structure that combines autoencoder, contrastive learning, or/and MLP projectors to generate more representative latent representations for the final clustering stage. However, existing deep contrastive multi-view clustering ignores two key points: (i) the latent representations projecting from one or more layers of MLP or new representations directly obtained from autoencoder fail to mine inherent relationship inner-view or cross-views; (ii) more existing frameworks only employ a one or dual-contrastive learning module, i.e., view- or/and category-oriented, which may result in the lack of communication between latent representations and clustering assignments. This paper proposes a new composite attention framework for contrastive multi-view clustering to address the above two challenges. Our method learns latent representations utilizing composite attention structure, i.e., Hierarchical Transformer for each view and Shared Attention for all views, rather than simple MLP. As a result, the learned representations can simultaneously preserve important features inside the view and balance the contributions across views. In addition, we add a new communication loss in our new dual contrastive framework. The common semantics will be brought into clustering assignments by pushing clustering assignments closer to the fused latent representations. Therefore, our method will provide a higher quality of clustering assignments for the segmentation problem of unlabeled multi-view data. The extensive experiments on several real data demonstrate that the proposed method can achieve superior performance over many state-of-the-art clustering algorithms, especially the significant improvement of an average of 10% on datasets Caltech and its subsets according to accuracy.

Improved smart energy-based routing approach for IoT networks in wireless sensor nodes

An emerging communication and network domain called the Internet of Things (IoT) uses nodes that function like actual real-world objects. Both moving and fixed nodes are possible for these. Maintaining node stability or improving inter-node connectivity is the main task for every IoT network. These networks are portable, leading to an overwhelming production of control overhead transmissions which can be reduced with the help of network clustering. In order to extend the network’s lifespan, it is therefore necessary to improve the power usage of sensor nodes. The goal of this research work is to improve network lifetime, increase throughput, reduce packet latency or packet loss, and even expand to addressing threatening nodes in order to cope with the issues of IoT-based sensor systems. The proposed Improved Smart Energy-Based Routing Protocol (ISERP), once combined with a wireless energy harvesting (WEH) unit, improves the node’s lifespan, the network’s quality of services (QoS) amid higher unequal congestion, and those aspects controlling its availability of energy. In the framework of distributed IoT, this protocol meets the standards of QoS through employing hardware-based link quality estimation and accomplishing very high efficiency in terms of energy in comparison with existing routing methods. The sink node chooses both the cluster head (CH) and cluster gateway (CG) by considering the estimation of its cost function. Simulation findings indicate that this proposed technique exhibits more network longevity along with lower consumption of energy compared with other existing cluster-based routing algorithms. Specifically, with the proposed ISERP, CHs use 33% less energy in comparison to existing techniques, and 40% of all nodes remain active until the end of the phase.

Identification and classification of the genomes of novel microviruses in poultry slaughterhouse.

Microviridae is a family of phages with circular ssDNA genomes and they are widely found in various environments and organisms. In this study, virome techniques were employed to explore potential members of Microviridae in a poultry slaughterhouse, leading to the identification of 98 novel and complete microvirus genomes. Using a similarity clustering network classification approach, these viruses were found to belong to at least 6 new subfamilies within Microviridae and 3 higher-level taxonomic units. Genome size, GC content and genome structure of these new taxa showed evident regularities, validating the rationality of our classification method. Our method can divide microviruses into about 45 additional detailed clusters, which may serve as a new standard for classifying Microviridae members. Furthermore, by addressing the scarcity of host information for microviruses, the current study significantly broadened their host range and discovered over 20 possible new hosts, including important pathogenic bacteria such as Helicobacter pylori and Vibrio cholerae, as well as different taxa demonstrated different host specificities. The findings of this study effectively expand the diversity of the Microviridae family, providing new insights for their classification and identification. Additionally, it offers a novel perspective for monitoring and controlling pathogenic microorganisms in poultry slaughterhouse environments.

Research on WSN clustering algorithm of firefly algorithm to optimize k-means

Aiming at the problem that uneven clustering of wireless sensor networks leads to large energy consumption differences between clusters and short network lifetime, this paper proposes a clustering routing algorithm of WSN based on the Firefly algorithm to optimize k-means (FAKMC). This paper mainly studies the clustering stage of the network, adopts the k-means algorithm for node partition, and uses the improved FA to optimize the sensitive problem of the initial clustering centre of k-means. Simulation results show that, compared to LEACH and k-means, the clustering result proposed in this paper is more uniform and can extend the network life cycle.

Resilience of finite clusters of carbon flux network under localized attack.

The investigation into the resilience of the carbon flux network regarding its capability to sustain the normal flow and transformation of carbon under extreme climatic events, pollutant emissions, biological invasions, and other factors, and the stability of connections between its nodes, has not yet been deeply studied. In this study, we developed carbon flux network models for various regional lands using complex networks, percolation theory, and introducing time delay effects using carbon flux daily data from 2000 to 2019 for three regions: China, the mainland United States, and Europe, to measure the resilience of finite clusters with sizes greater than or equal to s of the carbon flux network under localized attack. The analysis revealed that the carbon flux networks in different regions are characterized by a degree distribution consistent with the Poisson distribution. The carbon flux network demonstrated continuous phase transition behavior under localized attack. Interestingly, numerical simulation revealed a consistent relationship between the carbon flux network and the theoretical Erdős-Rényi network model. Moreover, the carbon flux network becomes more vulnerable as s increases. In addition, we discovered that there is a general scaling relationship of critical exponent δ≈-2 between the fraction of finite clusters and s. Therefore, investigating the resilience of carbon flux networks can enable us to predict and respond to the various risks and challenges, which will help policy designers formulate appropriate response strategies and enhance carbon flux systems' stability and resilience.

Symptom cluster study undergoing chemotherapy in breast cancer patients: Latent class analysis and contemporaneous network analysis

PurposeThis study aims to explore the subgroups and networks of symptom clusters in breast cancer patients undergoing chemotherapy, and to provide effective interventions for the core symptoms. Patients and methodsA cross-sectional survey was conducted at four comprehensive hospitals in Foshan City, China, from August to November 2023. A total of 292 participants completed the social determinants of health questionnaire, the numerical rating scale (NRS), the Pittsburgh sleep quality index (PSQI), the Chinese version of the cancer fatigue scale (CFS), and the hospital anxiety and depression Scale (HADS). Latent class analysis (LCA) was utilized to distinguish subgroups, and network analysis was utilized to identify core symptoms among different subgroups. ResultsBreast cancer patients undergoing chemotherapy exhibit symptoms were divided into two subgroups: the high burden group of symptoms (72.3%, Class 1) and the low burden group of symptoms (27.7%, Class 2). Education attainment, work status, family monthly income per capita, and daily sleep duration (hours) were associated with subgroup membership. “Panic feelings” (# HADS-A11) were the core symptom in both the full sample and Class 2, while “tension or pain” (# HADS-A1) was the core symptom in Class 1. ConclusionsThe core symptoms of fear, enjoyment, nervousness, and pain varied across subgroups of patients and could inform the current strategies for symptom management in breast cancer chemotherapy patients.

Planning and resource allocation of a hybrid IoT network using artificial intelligence

This paper introduces a pioneering hybrid topology tailored for Internet of Things (IoT) applications, integrating mesh and star wireless sensor configurations. This hybridized approach aims to optimize energy consumption efficiency while ensuring comprehensive network coverage for sensor deployment. The formulation of the network strategy is rooted in empirical data collected from real sensors deployed across two neighboring municipalities within the State of Espírito Santo, Brazil. Specifically, our analysis encompasses 380 strategically positioned sensors throughout Vitória city, all intended for connectivity to a central gateway located in Vila Velha. To establish mesh network clusters, we employed the k-Medoids algorithm for clustering fusion, while the GA with a binary solution was utilized to determine the star network points. In this approach, Dijkstra and genetic algorithms with real solutions are incorporated to facilitate efficient resource allocation within the mesh (utilizing ZigBee) and star (utilizing LoRa) networks. These resource allocation strategies are devised with the overarching objective of minimizing energy consumption. The findings of this investigation demonstrate that through the implementation of planning and resource allocation algorithms, we were able to effectively reduce the number of mesh networks and allocate resources to each designated end-point sensor.

An online transfer learning based multifactorial evolutionary algorithm for solving the clustered Steiner tree problem

The rapid increase in the number and computing power of devices interacting through the network over the past few years has led to a growing trend of decentralized clustering network architecture. The Clustered Steiner Tree Problem (CluSteiner) is a recently introduced NP-Hard problem and is assessed to be the core of efficient multicast routing in such networks. However, research on this problem in terms of theory and algorithmic design is still in its infancy as current solving approaches are limited in the literature. Lately, the Multifactorial Evolutionary Algorithm (MFEA) has been applied to solve several clustering-structure network design problems for the reason that these problems rarely exist independently and are often deployed concurrently in practice. To effectively transfer knowledge between problems while solving them simultaneously, this paper proposes an approach based on the novel data-driven MFEA-II to deal with the CluSteiner problem. In the proposal, an efficient encoding and a two-level decoding mechanism are introduced, combined with a module capable of online learning the amount of knowledge transferred between tasks that help the algorithm to explore potential regions of the search space. Experiments are undertaken on various test instances to evaluate the efficacy of the proposed algorithm. The empirical results indicate that our proposal can perform well on both metric and non-metric graphs with the statistical tests used to verify its efficiency over other baseline algorithms regarding the solution quality and convergence trend.

Perspectives of Perceived Organizational Support (POS) Theory Towards Individual Happiness

Introduction: Organizations must take a proactive stance when it comes to their workers' welfare and well-being to demonstrate their best commitment in this regard. Objective: This paper presents the descriptive data of organizational factors that contribute to staff happiness in one of the Malaysian Technical University Network (MTUN) clusters in Malaysia. Methods: This study was conducted at one of the Malaysian Technical University Network (MTUN) clusters in Malaysia. It involved 20 respondents comprising academicians who had served more than 11 years. The session at the Round Table Discussion was organized to discuss the issues that were highlighted in the survey data. Results and Conclusion: Analysis showed that the personal economy and personal appreciation dimension were among the highest contributors in determining employee happiness, followed by a support system, workstation environment, and trust. Furthermore, the discussion is related to the general aspect of Perceived Organizational Support Theory which supports the data analysis on individual happiness aspect in the institution of higher education. Originality/Value: By emphasizing the role of factors such as support systems, workstation environment, and trust in determining employee happiness, the study provides valuable insights into the broader dynamics of the workplace. Understanding how these factors influence employee satisfaction can inform strategies for improving organizational culture and employee engagement. Hence, this finding is related to Sustainable Development Goal (SDG) 3: Good Health and Well-being because it focuses on individual happiness, which is closely tied to mental well-being and overall health, both of which are central to SDG 3.