Search

This paper introduces a novel supervised classification method based on dynamic clustering (DC) and K-nearest neighbor (KNN) learning algorithms, denoted DC-KNN. The aim is to improve the accuracy of a classifier by using a DC method to discover the hidden patterns of the apriori groups of the training set. It provides a partitioning of each group into a predetermined number of subgroups. A new objective function is designed for the DC variant, based on a trade-off between the compactness and separation of all subgroups in the original groups. Moreover, the proposed DC method uses adaptive distances which assign a set of weights to the variables of each cluster, which depend on both their intra-cluster and inter-cluster structure. DC-KNN performs the minimization of a suitable objective function. Next, the KNN algorithm takes into account objects by assigning them to the label of subgroups. Furthermore, the classification step is performed according to two KNN competing algorithms. The proposed strategies have been evaluated using both synthetic data and widely used real datasets from public repositories. The achieved results have confirmed the effectiveness and robustness of the strategy in improving classification accuracy in comparison to alternative approaches. [ABSTRACT FROM AUTHOR]

In order to solve the problem that the unbalanced positional distribution of randomly screened cluster head nodes and unreasonable data transmission paths of forwarding nodes of the cluster routing protocol for wireless sensor networks will exacerbate the node's energy consumption and shorten the network's survival period, a multifactorial balanced dynamic cluster routing protocol based on improved social network search (ISNS) algorithm optimized fuzzy C-mean clustering (FCM) (multi-factor balanced dynamic clustering routing protocol, MD-LEACH) was proposed. Firstly, the Levy flight was introduced to improve the reverse elite learning strategy to enhance the global optimization capability of the social network search algorithm. Then, the ISNS-optimized fuzzy C-mean clustering algorithm was used to dynamically cluster the network nodes uniformly and balance the load of the network. Moreover, fuzzy reasoning was introduced to consider the energy factor and the location factor of nodes in the clusters in each cluster, and two cluster-head selection modes were designed to dynamically elect the cluster head to improve the Cluster head quality. In the stable transmission phase, the single hop was changed to the way of communication between cluster heads, and the improved ant colony algorithm was used to find the optimal data transmission path to improve the energy efficiency. Simulation results show that the algorithm can effectively improve the energy efficiency, balance the network load, and extend the network survival period. [ABSTRACT FROM AUTHOR]

Short text streams, such as social media comments, are continuously generated, making effective clustering methods essential for extracting valuable information. However, existing research fails to address the problem of topic concentration in clustering, which leads to multiple topics being confused in one cluster, making it challenging to summarize the center of clustering. To tackle this issue, this paper proposes a novel topic-enhanced clustering method called TEDM, based on the Dirichlet model. The method uses dynamic clustering, leveraging topic information to improve the sampling of documents and better cluster documents on the same topic. TEDM constructs a dynamic word relation graph to extract topic terms, which is updated with the stream of documents to cope with the dynamic changes in topics. Extensive experimental studies demonstrate that TEDM outperforms state-of-the-art works on multiple real datasets. [ABSTRACT FROM AUTHOR]

We introduce a new dynamic clustering method for multivariate panel data characterized by time-variation in cluster locations and shapes, cluster compositions, and possibly the number of clusters. To avoid overly frequent cluster switching (flickering), we extend standard cross-sectional clustering techniques with a penalty that shrinks observations toward the current center of their previous cluster assignment. This links consecutive cross-sections in the panel together, substantially reduces flickering, and enhances the economic interpretability of the outcome. We choose the shrinkage parameter in a data-driven way and study its misclassification properties theoretically as well as in several challenging simulation settings. The method is illustrated using a multivariate panel of four accounting ratios for 28 large European insurance firms between 2010 and 2020. [ABSTRACT FROM AUTHOR]

Persistently, urban regions grapple with the ongoing challenge of vehicular traffic, a predicament fueled by the incessant expansion of the population and the rise in the number of vehicles on the roads. The recurring challenge of vehicular congestion casts a negative influence on urban mobility, thereby diminishing the overall quality of life of residents. It is hypothesized that a dynamic clustering method of vehicle trajectory data can provide an accurate and up-to-date representation of real-time traffic behavior. To evaluate this hypothesis, data were collected from three different cities: San Francisco, Rome, and Guayaquil. A dynamic clustering algorithm was applied to identify traffic congestion patterns, and an indicator was applied to identify and evaluate the congestion conditions of the areas. The findings indicate a heightened level of precision and recall in congestion classification when contrasted with an approach relying on static cells. [ABSTRACT FROM AUTHOR]

The application of neural network models is becoming increasingly widespread.One of the most relevant aspects of neural networks is their generalization ability, which predicts situations that are not included in the training set. Research has shown that training set samples play a dominant role in the learning and training of back propagation (BP) neural network model, and the information contained in them directly affects the network performance. A large number of similar samples not only prolongs the training time of the model, but also leads to problems such as decreased network generalization ability. Based on this, this study focused on the impact of training set samples on the generalization ability of the BP neural network model. A model was constructed on this basis by introducing dynamic clustering methods to screen the training set samples with representative and typical features. The empirical results indicate that using dynamic clustering methods to screen samples can effectively remove redundant information from the training set, enhance the network generalization ability, and improve the model prediction accuracy.Moreover, for large datasets, it can effectively improve the model’s fitting degree and convergence speed, and enhance the “noise tolerance” performance of the BP neural network model.

Currently, supervised person re-identification (Re-ID) models trained on labeled datasets can achieve high recognition performance in the same data domain. However, accuracy drops dramatically when these models are directly applied to other unlabeled datasets or natural environments, due to a significant sample distribution gap between the two domains. Unsupervised Domain Adaptation (UDA) methods can solve this problem by fine-tuning the model on the target dataset with pseudo-labels generated by the clustering method. Yet, these methods are primarily aimed at the image-based person Re-ID domain. This is because the background noise and interference information are complex and changeable in the video scenarios, resulting in large intra-class distances and small inter-class spaces, which easily lead to noisy labels. Huge domain gap and noisy labels hinder clustering and training processes heavily in the video-based person Re-ID. To address the problem, we propose a novel UDA method via Dynamic Clustering and Co-segment Attentive Learning (DCCAL) for it. DCCAL includes a Dynamic Clustering (DC) module and a Co-segment Attentive Learning (CAL) module. The DC module is responsible for adaptively clustering pedestrians within different generation processes to alleviate noisy labels. On the other hand, the CAL module reduces the domain gap using a co-segmentation-based attention mechanism. Additionally, we introduce Kullback-Leibler (KL) divergence loss to reduce the distribution of features between two domains for better performance. Experimental results on two large-scale video-based person Re-ID datasets, MARS and DukeMTMC-VideoReID (DukeV), demonstrate exceptional precision performance. Our method outperforms state-of-the-art semi-supervised and unsupervised approaches by 1.1% in Rank-1 and 1.5% in mAP on DukeV, as well as 3.1% and 2.1% in Rank-1 and mAP on MARS, respectively.

In vehicular networks, diverse safety information can be shared among vehicles through internet connections. In vehicle-to-internet communications, vehicles on the road are wirelessly connected to different cloud networks, thereby accelerating safety information exchange. Onboard sensors acquire traffic-related information, and reliable intermediate nodes and network services, such as navigational facilities, allow to transmit safety information to distant target vehicles and stations. Using vehicle-to-network communications, we minimize delays and achieve high accuracy through consistent connectivity links. Our proposed approach uses intermediate nodes with two-hop separation to forward information. Target vehicle detection and routing of safety information are performed using machine learning algorithms. Compared with existing vehicle-to-internet solutions, our approach provides substantial improvements by reducing latency, packet drop, and overhead.

The evolution of agriculture towards a modern, intelligent system is crucial for achieving sustainable development and ensuring food security. In this context, leveraging the Internet of Things (IoT) stands as a pivotal strategy to enhance both crop quantity and quality while effectively managing natural resources such as water and fertilizer. Wireless sensor networks, the backbone of IoT-based smart agricultural infrastructure, gather ecosystem data and transmit them to sinks and drones. However, challenges persist, notably in network connectivity, energy consumption, and network lifetime, particularly when facing supernode and relay node failures. This paper introduces an innovative approach to address these challenges within heterogeneous wireless sensor network-based smart agriculture. The proposed solution comprises a novel connectivity management scheme and a dynamic clustering method facilitated by five distributed algorithms. The first and second algorithms focus on path collection, establishing connections between each node and m-supernodes via k-disjoint paths to ensure network robustness. The third and fourth algorithms provide sustained network connectivity during node and supernode failures by adjusting transmission powers and dynamically clustering agriculture sensors based on residual energy. In the fifth algorithm, an optimization algorithm is implemented on the dominating set problem to strategically position a subset of relay nodes as migration points for mobile supernodes to balance the network's energy depletion. The suggested solution demonstrates superior performance in addressing connectivity, failure tolerance, load balancing, and network lifetime, ensuring optimal agricultural outcomes. [ABSTRACT FROM AUTHOR]

Addressing sustainable mobility in urban areas has become a priority in today's society, given the growing population and increasing vehicular flow in these areas. Intelligent Transportation Systems have emerged as innovative and effective technological solutions for addressing these challenges. Research in this area has become crucial, as it contributes not only to improving mobility in urban areas but also to positively impacting the quality of life of their inhabitants. To address this, a dynamic clustering methodology for vehicular trajectory data is proposed which can provide an accurate representation of the traffic state. Data were collected for the city of San Francisco, a dynamic clustering algorithm was applied and then an indicator was applied to identify areas with traffic congestion. Several experiments were also conducted with different parameterizations of the forgetting factor of the clustering algorithm. We observed that there is an inverse relationship between forgetting and accuracy, and the tolerance allows for a flexible margin of error that allows for better results in precision. The results showed in terms of precision that the dynamic clustering methodology achieved high match rates compared to the congestion indicator applied to static cells. [ABSTRACT FROM AUTHOR]

In vehicular networks, diverse safety information can be shared among vehicles through internet connections. In vehicle‐to‐internet communications, vehicles on the road are wirelessly connected to different cloud networks, thereby accelerating safety information exchange. Onboard sensors acquire traffic‐related information, and reliable intermediate nodes and network services, such as navigational facilities, allow to transmit safety information to distant target vehicles and stations. Using vehicle‐to‐network communications, we minimize delays and achieve high accuracy through consistent connectivity links. Our proposed approach uses intermediate nodes with two‐hop separation to forward information. Target vehicle detection and routing of safety information are performed using machine learning algorithms. Compared with existing vehicle‐to‐internet solutions, our approach provides substantial improvements by reducing latency, packet drop, and overhead. [ABSTRACT FROM AUTHOR]

Point cloud segmentation is one of the most important tasks in LiDAR remote sensing with widespread scientific, industrial, and commercial applications. The research thereof has resulted in many breakthroughs in 3D object and scene understanding. Existing methods typically utilize hierarchical architectures for feature representation. However, the commonly used sampling and grouping methods in hierarchical networks are not only time-consuming but also limited to point-wise 3D coordinates, ignoring the local semantic homogeneity of point clusters. To address these issues, we propose a novel 3D point cloud representation network, called Dynamic Clustering Transformer Network (DCTNet). It has an encoder–decoder architecture, allowing for both local and global feature learning. Specifically, the encoder consists of a series of dynamic clustering-based Local Feature Aggregating (LFA) blocks and Transformer-based Global Feature Learning (GFL) blocks. In the LFA block, we propose novel semantic feature-based dynamic sampling and clustering methods, which enable the model to be aware of local semantic homogeneity for local feature aggregation. Furthermore, instead of traditional interpolation approaches, we propose a new semantic feature-guided upsampling method in the decoder for dense prediction. To our knowledge, DCTNet is the first work to introduce semantic information-based dynamic clustering into 3D Transformers. Extensive experiments on an object-based dataset (ShapeNet), and an airborne multispectral LiDAR dataset demonstrate the State-of-the-Art (SOTA) segmentation performance of DCTNet in terms of both accuracy and efficiency. Our code will be made publicly available.

Artificial intelligence has many applications in various industries, including agriculture. It can help overcome challenges by providing efficient solutions, especially in the early stages of development. When working with tree leaves to identify the type of disease, diseases often show up through changes in leaf color. Therefore, it is crucial to improve the color brightness before using them in intelligent agricultural systems. Color improvement should achieve a balance where no new colors appear, as this could interfere with accurate identification and diagnosis of the disease. This is considered one of the challenges in this field. This work proposes an effective model for olive disease diagnosis, consisting of five modules: image enhancement, feature extraction, clustering, and deep neural network. In image enhancement, noise reduction, balanced colors, and CLAHE are applied to LAB color space channels to improve image quality and visual stimulus. In feature extraction, raw images of olive leaves are processed through triple convolutional layers, max pooling operations, and flattening in the CNN convolutional phase. The classification process starts by dividing the data into clusters based on density, followed by the use of a deep neural network. The proposed model was tested on over 3200 olive leaf images and compared with two deep learning algorithms (VGG16 and Alexnet). The results of accuracy and loss rate show that the proposed model achieves (98%, 0.193), while VGG16 and Alexnet reach (96%, 0.432) and (95%, 1.74), respectively. The proposed model demonstrates a robust and effective approach for olive disease diagnosis that combines image enhancement techniques and deep learning-based classification to achieve accurate and reliable results. [ABSTRACT FROM AUTHOR]

Underwater wireless sensor networks possess considerable potential to monitor large and hostile underwater environments by reliably sensing, collecting, and forwarding data toward the surface sinks. Although the research community has made promising efforts, barriers such as continuous node mobility, longer delays, unavailability of location information, and energy limitations must be addressed. Taking this into account, this research aims to develop a hybrid and intelligent data collection scheme that considers node position and network characteristics during data forwarding. To accomplish the objective, the network is divided into two layers. The top layer, considered more dynamic, follows a hop-by-hop data forwarding scheme. The lower layer, experiencing stable water currents, follows a clustering-based data collection method. The proposed scheme, called Multilayer Dynamic Data Forwarding (MD2F), is suitable for large and deep underwater areas. MD2F is scalable as it uses a multi-sink architecture, while single or multiple autonomous underwater vehicles (AUVs) can be utilized depending on the area being monitored. Implementing hop-by-hop transmission and clustering-based data collection at different layers balances the network load, thereby increasing the network life. Results show that MD2F exhibits better performance when compared with Multilayer Cluster-based Energy Efficient (MLCEE) and Energy efficient and link reliable routing (E2LR) schemes, both are very close in working behavior. The results are encouraging in terms of delivery ratio, network throughput, and end-to-end delays. Alongside achieving these targets, the network also exhibits less energy consumption through load balancing.

The main objective of this study is to optimize the performance of a wireless sensor network (WSN) in terms of energy consumption and data routing delay using evolutionary metaheuristics. A WSN mobile sink-based routing method in which the sink moves to certain selected rendezvous nodes for data collection is used to address the hot-spot problem. However, there are two challenges, namely, clustering and mobile sink shortest trajectory traversal, which significantly affect the network energy consumption, lifetime, and delay. To achieve the goal of this study, first, a formal model is presented to solve the optimization problem of determining the optimal number of clusters and corresponding cluster heads, which are taken as rendezvous nodes accessed by the mobile sink considering the network energy consumption, intracluster communication, and transmission delay. Second, a discrete differential evolution algorithm is proposed to solve the formulated optimization problem. Third, an ant colony optimization-based algorithm is proposed to construct the shortest path for the mobile sink to traverse the selected rendezvous nodes. The experimental evaluations show that the proposed strategy significantly improves cluster head selection, balances the cluster members, increases the network lifetime by 54%, decreases the transmission delay by 63%, and reduces energy consumption by 47% compared to several routing strategies in the literature.

Green VANET is an emerging field of research that spurs interest in energy consumption management for the development of smart cities. It also presents a unique variety of research problems for developing trustworthy and scalable routing protocols as vehicles are susceptible to the restrictions of road geometry and the barriers which limit networking capabilities in urban environments. Clustering is a process of assembling vehicle nodes to create a powerful and effective network infrastructure. According to recent studies, clustering-based routing algorithms in green VANET may significantly improve networking effectiveness and lower infrastructure costs. However, sometimes the vehicle nodes are unaware of their OBU energy consumption, which causes network execution issues and topological alterations. In such instances, the energy consumption of onboard sensors becomes a major problem in the clustering-based routing protocol. Hence, this paper proposes a self-organizing map neural network (SOMNN) based dynamic clustering model to identify energy-efficient cluster head nodes from each cluster for vehicular data uploading and downloading applications. The simulation results demonstrate that the proposed model solves network lifetime issues and provides superior network effectiveness with enhanced communication stability. The suggested dynamic clustering model reduces network energy consumption by 26% and 18% in comparison to k - means (KM) and fuzzy c - means (FCM) based clustering model. [ABSTRACT FROM AUTHOR]

Load scheduling and peak demand shaving are two critical aspects of the utility grid operation that help both the grid operators as well as end-users. This paper proposes a two-stage community home energy management system for microgrids. The first stage deals with the dynamic clustered community load scheduling scheme. Comparatively flatter power demand was attained using particle swarm optimization (PSO) incorporating user-defined constraints. The new arising or remaining peaks as a consequence of consumer constraints are catered to in the second stage. The second stage proposes a rule-based peak shaving management method for the photovoltaic (PV) systems that are connected to the grid and the battery energy storage systems. The proposed technique determines the dynamic demand and feed-in limits based on the estimations of the upcoming day’s load demand and PV power profiles. Also, the study presents an optimal rule-based management technique for peak shaving of utility grid power that sets the charge/discharge day ahead schedules of the battery. For peak energy minimization, PSO is used to calculate the optimal inputs needed for implementing the appropriate rule-based management strategy. MATLAB is used to test the proposed method for different PV power and load demand patterns, thus, achieving an average improvement of 8.5%.

The relevance of the study lies in the need for a scientific search for the possibilities of using new types of Big data in studies of the population vulnerability to solve practical problems of improving the safety of urban spaces from natural and man-made hazards. The object of the study is the administrative districts of Moscow; the subject is the temporal patterns of vulnerability of their population to potential natural and man-made hazards. The research question of the study is to develop a typology of Moscow districts and further assess this sustainability in terms of the population vulnerability to natural and man-made hazards. To achieve this research question, a set of tasks was solved: 1. Processing of the mobile operators’ data array and further construction of a continuous graph of the Moscow population dynamics in 2019 (with a time cycle of 30 minutes, over 36 million measurements in more than 7 thousand time slices); 2. Empirical justification of natural temporal boundaries of daily, weekly, seasonal cycles of population dynamics in Moscow districts; 3. Justification of key factors and parameters of urban population vulnerability; 4. Development and approbation of the dynamic clustering method of Moscow districts using selected variables and periods. The study is based on the impersonal mobile operators’ data on the locations of subscribers for 2019, provided by the Department of Information Technologies of the Moscow city. The method of dynamic cluster analysis is used. Four particular clusterings were obtained that characterize the “behavior” of the settlement system in the main intervals of social time (weekdays and weekends of the cold and warm seasons). Сluster stability matrix allows to identify which of the districts retain their properties during the period under review, and which are characterized by instability of considered indicators of population vulnerability. Depending on the stability of the position of the districts in a particular cluster, “stable”, “conditionally stable” and “nomadic” types of districts were identified. The study showed that the first two types include spatial-settlement structures that are stable in time with approximately the same level of population vulnerability during the year, while the third type requires a special differentiated approach to the development of measures to protect the population from natural and man-made emergencies. Calculations have shown that “nomadic” type of districts concentrate on average from 2.2 million people in the summer season to 3 million people in the winter season, that is, a very significant share of the entire population of the capital.

MANETs aredecentralized network that involves mobile nodes. As the overall network is mobile and has no centralization, network management, routing, and security become very challenging. Though many works have been presented, still there is a lack in organizing the network due to unauthorized access, centralized security schemes, and the dynamic nature of the nodes. This paper proposed a novel Blockchain-assisted Secure Routing (Block-Sec) protocol for MANETs. All mobile nodes are authenticated by Distributed One-Time Passcode (DOT) based authorization scheme. All authorized nodes are segregated into multiple clusters based on Weight based Dynamic Clustering (WDC) algorithm in which multiple metrics are considered in clustering and re-clustering processes. After cluster formation, each cluster is elected with optimal Cluster Head (CH) by Strawberry Optimization (SBO) algorithm with a new objective function. After cluster formation, the optimal route is selected by Fast Neural Net-assisted Fuzzy (FNNF) algorithm by combining multiple variables. Data transmission is secured by Efficient Elliptic Curve (E2C2) algorithm. With the combined algorithms, the proposed approach obtainedimproved efficiency in packet delivery ratio (PDR), throughput, time analysis, and security level. [ABSTRACT FROM AUTHOR]

To deal with the insufficient feedbacks and dynamics of individual arrival and item popularity in online recommender, collaborative multi-armed bandit (MAB) schemes intentionally utilize the explicitly known or implicitly inferred social relationships among individuals to collaboratively recommend. Especially, without assuming the social relationships among individuals given, the dynamic cluster of bandits simultaneously infers the relationships, and recommends items through using the inferred relationships in multi-round interactive steps. However, the existed clustering bandit algorithms have two weakpoints: first they either fix the number of clusters in advance, or assign two individuals into the same cluster if there exists a path between two users in graph structure, which may lead to the wrongly cluster users. Second, they usually exploit only the cluster's accumulated parameters of cluster as the inferred preference of individual in the cluster, which can't fully accurately learn individual's latent preference. To address issues above, we propose new clustering MAB based online recommendation methods, ADCB and ADCB+, based on adaptively splitting and merging clusters, which incrementally enforce both user-level re-assignment and cluster-level re-adjustment in recommendation rounds to efficiently and effectively learn the individuals' preferences and their clustering structure. Especially, the proposed ADCB+ method further exploits both the accumulated cluster preference parameters and each individual's personalized feature through the adaptively weighting of the two influences according to the number of user interactions. The experiments on three real online rating datasets (i.e., MovieLens-2k, Delicious-2k, LastFM-2k) consistently show that, in terms of the cumulative reward over recommendation rounds, and the average Click-Through-Rate, our proposed ADCB and ADCB+ schemes outperform than some existing dynamic clustering based online recommendation methods. [ABSTRACT FROM AUTHOR]

Emergency event monitoring is a hot topic in wireless sensor networks (WSNs). Benefiting from the progress of Micro-Electro-Mechanical System (MEMS) technology, it is possible to process emergency events locally by using the computing capacities of redundant nodes in large-scale WSNs. However, it is challenging to design a resource scheduling and computation offloading strategy for a large number of nodes in an event-driven dynamic environment. In this paper, focusing on cooperative computing with a large number of nodes, we propose a set of solutions, including dynamic clustering, inter-cluster task assignment and intra-cluster one-to-multiple cooperative computing. Firstly, an equal-size K-means clustering algorithm is proposed, which activates the nodes around event location and then divides active nodes into several clusters. Then, through inter-cluster task assignment, every computation task of events is alternately assigned to the cluster heads. Next, in order to make each cluster efficiently complete the computation tasks within the deadline, a Deep Deterministic Policy Gradient (DDPG)-based intra-cluster one-to-multiple cooperative computing algorithm is proposed to obtain a computation offloading strategy. Simulation studies show that the performance of the proposed algorithm is close to that of the exhaustive algorithm and better than other classical algorithms and the Deep Q Network (DQN) algorithm. [ABSTRACT FROM AUTHOR]

Underwater wireless sensor networks possess considerable potential to monitor large and hostile underwater environments by reliably sensing, collecting, and forwarding data toward the surface sinks. Although the research community has made promising efforts, barriers such as continuous node mobility, longer delays, unavailability of location information, and energy limitations must be addressed. Taking this into account, this research aims to develop a hybrid and intelligent data collection scheme that considers node position and network characteristics during data forwarding. To accomplish the objective, the network is divided into two layers. The top layer, considered more dynamic, follows a hop-by-hop data forwarding scheme. The lower layer, experiencing stable water currents, follows a clustering-based data collection method. The proposed scheme, called Multilayer Dynamic Data Forwarding (MD2F), is suitable for large and deep underwater areas. MD2F is scalable as it uses a multi-sink architecture, while single or multiple autonomous underwater vehicles (AUVs) can be utilized depending on the area being monitored. Implementing hop-by-hop transmission and clustering-based data collection at different layers balances the network load, thereby increasing the network life. Results show that MD2F exhibits better performance when compared with Multilayer Cluster-based Energy Efficient (MLCEE) and Energy efficient and link reliable routing (E2LR) schemes, both are very close in working behavior. The results are encouraging in terms of delivery ratio, network throughput, and end-to-end delays. Alongside achieving these targets, the network also exhibits less energy consumption through load balancing. [ABSTRACT FROM AUTHOR]

3D non-rigid shape correspondence is significant but challenging in computer graphics, computer vision, and related fields. Although some deep neural networks have achieved encouraging results in shape correspondence, due to the complexity of the local deformation of non-rigid shapes, the ability of these networks to identify the spatial changes of objects is still insufficient. In this paper, we design a Component-aware Capsule Graph Network (CA-CGNet) to further address the features of embedding space based on the component constraints. Specifically, the dynamic clustering strategy is used to classify the features of patches produced by over-segmentation in order to further reduce noise interference. Moreover, aiming at the problem that existing routing ignores the embedding relationship between capsules, we propose a component-aware capsule graph routing to fully describe the relationship between capsules, which regards capsules as nodes in the graph network and constrains nodes through component information. Then, a knowledge distillation strategy is introduced to improve the convergence speed of the network by decreasing the parameters while maintaining accuracy. Finally, a component pair constraint is added to the functional map, and the component-based semantic loss function is proposed, which can compute isomeric in both direct and symmetric directions. The experimental results show that CA-CGNet improves by 10.21% compared with other methods, indicating the accuracy, generalization, and efficiency of our method on the FAUST, SCAPE, TOSCA, and KIDS datasets. [ABSTRACT FROM AUTHOR]

On-demand, automatic resource delivery in a transparent manner to users is a remarkable feature offered by the cloud computing environment. User demands are met by dynamically provisioning the cloud resources. Incidental failures during task execution in cloud could be attributed to a variety of reasons. Such failures bring down the cloud performance. Recently, a variety of intelligent task scheduling algorithms have been demonstrated to address several issues in cloud scheduling. Most of these algorithms neglect the fault tolerance criterion, which if addressed appropriately could contribute to better cloud performance. In this research, we had proposed a Dynamic Clustering Cuckoo Whale Optimization Algorithm (DCCWOA) for carrying out efficient scheduling in the cloud by paying equal attention to the fault tolerance parameter. The proposed fault tolerance aware algorithm addresses the scheduling of tasks by maintaining a tab on the currently available resources such that unfortunate failures of autonomous tasks get effectively addressed leading to reduced failures. The performance of the proposed fault tolerance aware DCCWOA has been compared with Ant Colony Optimization algorithm (ACO), Genetic Algorithm (GA) and League Championship Algorithm (LCA)with respect to makespan, failure ratio and failure slowdown parameters under three different scenarios, where in each scenario the number of tasks were appropriately varied. It has been found that the proposed DCCWOA had produced an improvement of 58.19%, 19.88% and 29.32% under scenario 1 for makespan, failure ratio and failure slowdown parameters respectively when compared to ACO, GA and LCA algorithms respectively. Detailed experimental results for scenarios 1, 2 and 3 had been presented in the results section of this article. Results obtained prove the efficacy of the proposed algorithm in overcoming the faults and increasing the scheduling performance of the cloud with respect to the failure rate. [ABSTRACT FROM AUTHOR]

Persistently, urban regions grapple with the ongoing challenge of vehicular traffic, a predicament fueled by the incessant expansion of the population and the rise in the number of vehicles on the roads. The recurring challenge of vehicular congestion casts a negative influence on urban mobility, thereby diminishing the overall quality of life of residents. It is hypothesized that a dynamic clustering method of vehicle trajectory data can provide an accurate and up-to-date representation of real-time traffic behavior. To evaluate this hypothesis, data were collected from three different cities: San Francisco, Rome, and Guayaquil. A dynamic clustering algorithm was applied to identify traffic congestion patterns, and an indicator was applied to identify and evaluate the congestion conditions of the areas. The findings indicate a heightened level of precision and recall in congestion classification when contrasted with an approach relying on static cells.

The evolution of agriculture towards a modern, intelligent system is crucial for achieving sustainable development and ensuring food security. In this context, leveraging the Internet of Things (IoT) stands as a pivotal strategy to enhance both crop quantity and quality while effectively managing natural resources such as water and fertilizer. Wireless sensor networks, the backbone of IoT-based smart agricultural infrastructure, gather ecosystem data and transmit them to sinks and drones. However, challenges persist, notably in network connectivity, energy consumption, and network lifetime, particularly when facing supernode and relay node failures. This paper introduces an innovative approach to address these challenges within heterogeneous wireless sensor network-based smart agriculture. The proposed solution comprises a novel connectivity management scheme and a dynamic clustering method facilitated by five distributed algorithms. The first and second algorithms focus on path collection, establishing connections between each node and m-supernodes via k-disjoint paths to ensure network robustness. The third and fourth algorithms provide sustained network connectivity during node and supernode failures by adjusting transmission powers and dynamically clustering agriculture sensors based on residual energy. In the fifth algorithm, an optimization algorithm is implemented on the dominating set problem to strategically position a subset of relay nodes as migration points for mobile supernodes to balance the network’s energy depletion. The suggested solution demonstrates superior performance in addressing connectivity, failure tolerance, load balancing, and network lifetime, ensuring optimal agricultural outcomes.

Energy conservation is one of the main challenges hindering the advancement of Wireless Sensor Networks (WSNs) in the Internet of Things (IoT). Although various approaches including probabilistic, deterministic, and mixed model based solutions have been proposed to solve the issue of energy conservation in cluster based WSNs, most of them do not consider the node’s heterogeneity in terms of energy as well as transmission rate during the creation or selection of clusters. In this paper, a new Fairness Aware Clustering Scheme (FACS) for heterogeneous WSN is proposed, where we consider heterogeneity of the WS N in terms of transmission rate and energy. In contrast to existing schemes, the proposed FACS provides a set of polices that take care of round operations and improve the network performance and lifetime. In FACS, fuzzy logic is employed to benefit from the heterogeneity of the nodes and to determine the utility of each node. Through extensive simulation, we show that FACS performs efficiently and it outperforms the existing state-of-the-art approaches in terms of network lifespan, energy, and fairness.

To optimize graded lattice structures with spatially graded geometrical characteristics and promote the rational distribution of finite materials in macrostructure, a two-scale concurrent topology optimization method combining dynamic clustering is proposed. Dynamic clustering strategy is developed to group microstructures, which means that the clustering pattern will be updated according to the current strain energy during the whole iterations and obtains a more reasonable microstructures’ distribution than the static partition. Further, rotational degree of freedom is introduced to make sure that the microstructures are oriented along principal stress directions, so as to obtain a more reasonable load transfer path. Numerical examples show that, compared with the traditional static partition method, this method can more effectively distribute materials, make full use of the anisotropic of lattice structures, and improve structural performance. [ABSTRACT FROM AUTHOR]

Copyright of Frontiers of Information Technology & Electronic Engineering is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

In recent years, topology optimization of periodic structures has become an effective approach to generating efficient designs that meet a variety of practical considerations, including manufacturability, transportability, replaceability, and ease of assembly. Traditional periodic structural optimization typically restricts designs to a uniform assembly configuration utilizing only one type of unit cell. This study proposed a novel clustering-based approach for periodic structural optimization, which allows variable orientations of individual unit cells. A dynamic k -means clustering strategy is introduced to categorize all unit cells into distinct groups and gradually eliminate less efficient unit cells from the optimized design. Meanwhile, a novel technique is introduced to identify and select more efficient orientations of unit cells during the optimization process. Several numerical examples are presented to demonstrate the effectiveness of the proposed approach. The results show that periodic structures with clustered oriented unit cells can significantly outperform their traditional periodic counterparts. This study not only incorporates assembly flexibility into periodic topology optimization but also utilizes multiple types of unit cells in a design, thereby further enhancing its structural performance. • A periodic optimization algorithm is developed for clustered oriented unit cells. • A novel technique is proposed to select more efficient unit cell orientations. • A dynamic clustering strategy incorporates multiple unit cell types in the design. • Structurally efficient designs with oriented unit cells can be easily obtained. [ABSTRACT FROM AUTHOR]

The advent of big data related to spatial position knowledge, called geospatial big data, provides us with opportunities to recognize the urban environment. Existing database processing methods are inadequate to rapidly provide reliable results in a geospatial big data context due to the need for defining approximation “measures” and the increasing execution time for the queries. The clustering method yields the functional effects. How to scale and accelerate clustering algorithms while maintaining high clustering efficiency, on the other hand, remains a significant challenge. The paper’s primary contribution is the introduction of a modified hierarchical distributed k-medoid clustering method that is specific to spatial query analysis for big data. To improve the efficiency of the k-medoid algorithm and obtain more precise clusters, the suggested model utilizes the Fuzzy k-Medoids method to overcome outliers in the spatial data set and to deal with data uncertainty. The method is complex in nature since it is not predicated on the number of right clusters. The proposed model is divided into two phases: the first step creates local clusters based on a portion of the entire dataset; this stage makes extensive use of the parallelism paradigm provided by the Apache Spark framework; and the second phase aggregates the local clusters to produce compact and reliable final clusters. The proposed model greatly reduces the amount of knowledge shared during the aggregation process and automatically produces the appropriate number of clusters based on the dataset characteristics. The results show that the proposed model outperforms the traditional K-medoids in terms of accuracy of obtained centers in big data applications.

Despite the decisive contribution of intelligent transport systems in road safety, they also open new vulnerabilities to cyber-attacks, particularly vehicle position-linked attacks. For that reason, centralized systems are becoming increasingly vulnerable to the growth of the connected-vehicle fleets as it becomes more challenging to revoke certificates in real-time. We have proposed a new method that integrates a decentralized, collaborative system to meet these challenges. This method efficiently allows Blockchain integration for vehicular network’s cyber security by dynamically creating communities to revoke malicious vehicles in real-time. This article presents analytical models of the system of real-time revoking certificates and examines our solution’s impact on two important types of attacks in V2X communications, Sybil and the faking position attacks. Our experiments using real V2X hardware demonstrated the feasibility and benefits of real-time revocation via vehicle communities. The results were obtained from consensus implementation in a vehicular network comprising three communicating vehicles and a single roadside unit. In parallel, simulations showed feasibility in large-scale communications. As a result, the exposure and detection times of our solution meet real-time requirements.

A new meta-heuristic algorithm called like-attracts-like optimizer (LALO) is proposed in this article. It is inspired by the fact that an excellent person (i.e., a high-quality solution) easily attracts like-minded people to approach him or her. This LALO algorithm is an important inspiration for video robotics cluster control. First, the searching individuals are dynamically divided into multiple clusters by a growing neural gas network according to their positions, in which the topological relations between different clusters can also be determined. Second, each individual will approach a better individual from its superordinate cluster and the adjacent clusters. The performance of LALO is evaluated based on unimodal benchmark functions compared with various well-known meta-heuristic algorithms, which reveals that it is competitive for some optimizations.

Open-pit mine networking is one of the basic key technologies of intelligent mines. However, the complex communication environment in open-pit mines and the large mobility difference of various heterogeneous nodes lead to unstable communication links. In order to solve this problem, a heterogeneous clustering networking protocol (HCNP) is proposed. HCNP supports the access of static solar LoRa relay equipment at any time, and can be flexibly arranged with the change of mining depth. HCNP uses the shortest path tree to form a static backbone network from static devices and static relay devices to base stations. HCNP divides the priority according to whether the node moves, whether the energy is limited, the number of hops and the distance to the base station. The cluster head selection and clustering process are completed based on the different priorities of the node, and a heterogeneous networking mode of static backbone network and dynamic clustering is formed so as to realize the reliable forwarding of data. The simulation results show that compared with Ad hoc on-demand distance vector routing (AODV), optimized link state routing protocol (OLSR), greedy peripheral stateless routing protocol (GPSR), and node location-based 3D geographical routing protocol (AB3D), HCNP has higher successful transmission rate, lower end-to-end delay and lower energy consumption. When the simulation time and the number of nodes change, the successful transmission rate is always higher than 83%, the end-to-end delay is lower than 20 ms, and the energy consumption does not exceed 0.66 J. This shows that the static backbone network and the clustering strategy considering node priority are suitable for 3D open-pit mines with complex environment, and meet the requirements of dynamic adaptability of mine equipment and personnel number.

Group decision support systems (GDSSs) have been widely studied over the recent decades. The Web-based group decision support systems appeared to support the group decision-making process by creating the conditions for it to be effective, allowing the management and participation in the process to be carried out from any place and at any time. In GDSS, argumentation is ideal, since it makes it easier to use justifications and explanations in interactions between decision-makers so they can sustain their opinions. Aspect-based sentiment analysis (ABSA) intends to classify opinions at the aspect level and identify the elements of an opinion. Intelligent reports for GDSS provide decision makers with accurate information about each decision-making round. Applying ABSA techniques to group decision making context results in the automatic identification of alternatives and criteria, for instance. This automatic identification is essential to reduce the time decision makers take to step themselves up on group decision support systems and to offer them various insights and knowledge on the discussion they are participating in. In this work, we propose and implement a methodology that uses an unsupervised technique and clustering to group arguments on topics around a specific alternative, for example, or a discussion comparing two alternatives. We experimented with several combinations of word embedding, dimensionality reduction techniques, and different clustering algorithms to achieve the best approach. The best method consisted of applying the KMeans++ clustering technique, using SBERT as a word embedder with UMAP dimensionality reduction. These experiments achieved a silhouette score of 0.63 with eight clusters on the baseball dataset, which wielded good cluster results based on their manual review and word clouds. We obtained a silhouette score of 0.59 with 16 clusters on the car brand dataset, which we used as an approach validation dataset. With the results of this work, intelligent reports for GDSS become even more helpful, since they can dynamically organize the conversations taking place by grouping them on the arguments used. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Science & Technology (JCS&T) is the property of Journal of Computer Science & Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Prior study suggests that VANET has two types of communications: Vehicle to Vehicle (V2V) and Vehicle to Infrastructure (V2I) communications. V2V is very important and ensures cooperative communications between vehicles and safety measures. It is also defined as Inter-Vehicle Communication (IVC).The communication is based on clustering the nodes to transmit the data from vehicle to vehicle. The overhead and stability are considered as main challenges that need to be addressed during vehicle intersections. In this paper, a novel self-adaptable Angular based k-medoid Clustering Scheme (SAACS) is proposed to form flexible clusters. The clusters are formed by estimating the road length and transmission ranges to minimize the network delay. And the Cluster Head (CH) is elected from a novel performance metric, 'cosine-based node uncoupling frequency,' that finds the best nodes irrespective of their current network statistics. The parametric analysis varies according to the number of vehicular nodes with the transmission range. The experimental results have proven that the proposed technique serves better in comparison to existing approaches such as Cluster Head Lifetime (CHL), Cluster Member Lifetime (CML), Cluster Number (CL), Cluster Overhead (CO), Packet Loss Ratio (PLR) and Average Packet Delay (APD). CHL is enhanced 40% as compare to Real-Time Vehicular Communication (RTVC), Efficient Cluster Head Selection (ECHS) whereas CML is 50% better than RTVC and ECHS. Packet loss ratio and overhead is 45% better in our proposed algorithm than RTVC and ECHS. It is observed from the results that the incorporation of cosine-based node uncoupling frequency has minimized the incongruity between vehicular nodes placed in dense and sparse zones of highways. [ABSTRACT FROM AUTHOR]

The potential of intelligent reflecting surfaces (IRSs) is investigated as a promising technique for enhancing the energy efficiency of wireless networks. Specifically, the IRS enables passive beamsteering by employing many low-cost individually controllable reflect elements. The resulting change of the channel state, however, not only increases the signal quality but also the interference at the users. To counteract this negative side effect, we employ rate splitting (RS), which inherently is able to mitigate the impact of interference. We facilitate practical implementation by considering a Cloud Radio Access Network (C-RAN) at the cost of finite fronthaul-link capacities, which necessitate the allocation of sensible user-centric clusters to ensure energy-efficient transmissions. Dynamic methods for RS and the user clustering are proposed to account for the interdependencies of the individual techniques. Numerical results show that the dynamic RS method establishes synergistic benefits between RS and the IRS. Additionally, the dynamic user clustering and the IRS cooperate synergistically, reflected by increased individual gains for the dynamic clustering. Interestingly, with an increasing fronthaul capacity, the gain of the dynamic user clustering decreases, while the gain of the dynamic RS method increases. Around the resulting intersection, both methods affect the system concurrently, improving the energy efficiency drastically. [ABSTRACT FROM AUTHOR]