Adaptive Clustering Research Articles

Substation clustering based on improved KFCM algorithm with adaptive optimal clustering number selection

The premise and basis of load modeling are substation load composition inquiries and cluster analyses. However, the traditional kernel fuzzy C-means (KFCM) algorithm is limited by artificial clustering number selection and its convergence to local optimal solutions. To overcome these limitations, an improved KFCM algorithm with adaptive optimal clustering number selection is proposed in this paper. This algorithm optimizes the KFCM algorithm by combining the powerful global search ability of genetic algorithm and the robust local search ability of simulated annealing algorithm. The improved KFCM algorithm adaptively determines the ideal number of clusters using the clustering evaluation index ratio. Compared with the traditional KFCM algorithm, the enhanced KFCM algorithm has robust clustering and comprehensive abilities, enabling the efficient convergence to the global optimal solution

Proposed energy efficient clustering and routing for wireless sensor network

Wireless sensor network (WSN) is considered a growing research field that includes numerous sensor nodes used to gather, process, and broadcast information. Energy efficiency is considered one of the challenging tasks in the WSN. The clustering and routing are considered capable approaches to solve the issues of energy efficiency and enhance the network’s lifetime. In this research, the multi-objective-energy based black widow optimization algorithm (M-EBWOA) is proposed to perform the cluster-based routing over the WSN. The M-EBWOA-based optimal cluster head discovery is used to assure an energy-aware routing over the WSN. The main goal of this M-EBWOA is to minimize the energy consumed by the nodes while improving the data delivery of the WSN. The performance of the M-EBWOA is analyzed as alive and dead nodes, dissipated energy, packets sent to base station, and life expectancy. The existing research such as low-energy adaptive clustering hierarchy (LEACH), hybrid grey wolf optimizer-based sunflower optimization (HGWSFO), genetic algorithm-particle swarm optimization (GA-PSO), and energy-centric multi-objective Salp Swarm algorithm (ECMOSSA) are used to evaluate the efficiency of M-EBWOA. The alive nodes of the M-EBWOA are 100 for 2,500 rounds, which is higher than the LEACH, HGWSFO, GA-PSO, and ECMOSSA.

Improved Adaptive Finch Clustering Sonar Segmentation Algorithm Based on Data Distribution and Posterior Probability

This study proposes a side-scan sonar target detection technique for CPU or low-performance GPU to meet the requirement of underwater target detection. To rectify the gray distribution of the original side scan sonar data, enhance picture segmentation, and supply the data distribution probability for the clustering algorithm, the methodology uses a classic image processing technique that is GPU-friendly. The modified adaptive Finch clustering technique is used to segment the image and remove image voids after assessing the processed image attributes. The posterior information is then used to apply a classification label to each pixel. The characteristics of the connected region are analyzed in the data playback of the Tuandao experiment in accordance with the imaging principle of side-scan sonar and the original shape and size characteristics of the target. The predicted target results are combined with the AUV navigation information to obtain the predicted target longitude and latitude information, which is then sent to the AUV master control system to guide the next plan. The Jiaozhou Bay sea test results demonstrate that the traditional target detection algorithm put forth in this paper can be integrated into a low-performance GPU to detect targets and locate them. The detection accuracy and speed exhibit strong performance, and real-time autonomous sonar detection is made possible.

Evolution Characteristics and Causes—An Analysis of Urban Catering Cluster Spatial Structure

Studying the development characteristics of the urban catering industry holds significant importance for understanding the spatial patterns of cities. In this manuscript, according to the characteristics of the distribution of catering points and based on catering point of interest (POI) data of 106 cities in China in 2016 and 2022, we propose the Natural Nearest Neighbor Single Branch Model (NNSBM) to identify catering points by adaptive clustering, which improves the efficiency of identifying catering clusters. Subsequently, a catering spatial structure division model is constructed to classify the spatial structure of catering clusters into 3 major categories and 17 subcategories, and the evolution pattern of urban catering clusters is analyzed. In addition, based on the population density raster data, a bivariate spatial autocorrelation model is employed to analyze the complex relationship between the distribution of urban catering clusters and population density, revealing the distinctive characteristics of urban catering cluster evolution. The results showed that (1) In the initial stage of catering cluster formation, catering activities tend to gather first in a specific area of the city, giving rise to the main catering cluster. However, as the catering industry progresses, the phenomenon of “central fading” occurs within the main catering cluster. (2) The overall trend of the catering spatial structure of most cities showed an evolution toward low primacy–high concentration (Lp-Hc), and cities at different stages of catering capacity exhibited different evolution characteristics of catering clusters. (3) The influence of population density on catering distribution was staged, with a varying impact on cities with different types of catering spatial structures.

Navigation Line Extraction Method for Broad-Leaved Plants in the Multi-Period Environments of the High-Ridge Cultivation Mode

Navigation line extraction is critical for precision agriculture and automatic navigation. A novel method for extracting navigation lines based on machine vision is proposed herein using a straight line detected based on a high-ridge crop row. Aiming at the low-level automation of machines in field environments of a high-ridge cultivation mode for broad-leaved plants, a navigation line extraction method suitable for multiple periods and with high timeliness is designed. The method comprises four sequentially linked phases: image segmentation, feature point extraction, navigation line calculation, and dynamic segmentation horizontal strip number feedback. The a* component of the CIE-Lab colour space is extracted to preliminarily extract the crop row features. The OTSU algorithm is combined with morphological processing to completely separate the crop rows and backgrounds. The crop row feature points are extracted using an improved isometric segmented vertical projection method. While calculating the navigation lines, an adaptive clustering method is used to cluster the adjacent feature points. A dynamic segmentation point clustering method is used to determine the final clustering feature point sets, and the feature point sets are optimised using lateral distance and point line distance methods. In the optimisation process, a linear regression method based on the Huber loss function is used to fit the optimised feature point set to obtain the crop row centreline, and the navigation line is calculated according to the two crop lines. Finally, before entering the next frame processing process, a feedback mechanism to calculate a number of horizontal strips for the next frame is introduced to improve the ability of the algorithm to adapt to multiple periods. The experimental results show that the proposed method can meet the efficiency requirements for visual navigation. The average time for the image processing of four samples is 38.53 ms. Compared with the least squares method, the proposed method can adapt to a longer growth period of crops.

Review study about Hierarchical Routing Protocol in wireless sensor networks

Wireless Sensor Networks (WSNs) are crucial in various applications, including smart homes, military surveillance, health monitoring, and environmental sensing. These networks consist of distributed sensor nodes that collect and transmit environmental data to a central location for processing. However, WSNs face challenges such as limited energy resources, scalability, complexity, security, fault tolerance, and quality of service. Researchers have proposed different routing protocols tailored to WSN characteristics to address these issues. This article comprehensively overviews WSN design issues, focusing on routing protocol classification and comparative analysis. Among the hierarchical protocols, LEACH (Low-Energy Adaptive Clustering Hierarchy) is highlighted as a prominent clustering-based protocol that reduces energy consumption by forming clusters and utilizing cluster heads for data aggregation. Each addresses specific challenges and enhances network performance. The comparative analysis evaluates the protocols based on energy efficiency, network lifetime, data delivery, scalability, and security. The findings help researchers, practitioners, and network designers select appropriate routing protocols for their specific WSN applications. This article offers valuable insights into WSN design issues and presents a comprehensive classification and comparison of routing protocols. By understanding the strengths and limitations of different protocols.

Energy management and balance strategy for underwater glider fleet with VMD-SSA-LSSVM model and central type double-level framework

The management of energy has gained enormous attention as an efficient way to prolong the endurance of unmanned vehicle fleet. During ocean observation and exploration, the variation of the equipped sensor will cause uneven energy consumption for different underwater gliders (UGs), which will lead to the premature accomplishment of the planned mission. Therefore, a novel energy management and balance strategy for the UG fleet is proposed in this paper. Firstly, a new VMD-SSA-LSSVM model is presented to predict the energy consumption of UG based on sea trial data acquired in the northern South China Sea. Besides, the prediction results are compared with those of other benchmark methods and proved their effectiveness. Afterwards, a novel central type double-level framework is established based on low energy adaptive clustering hierarchy protocol to balance the energy of UG fleet, and double-level managing strategies are provided for minor or major energy adjustment based on the energy consumption model. Then the simulation is carried out to verify the effect on different sizes of UG fleet with energy balancing strategy, and the observation efficiency is defined and quantified by many indexes comprehensively. Finally, the simulated results with real data of UG fleet are given, verifying that the energy balance strategy could prolong the endurance by about 4.6% and improve the observation efficiency by about 7%.

3D LIDAR SEGMENTATION BASED ON EUCLIDEAN CLUSTERING FOR EMBEDDED SYSTEM

This paper presents a method for 3D LiDAR segmentation based on Euclidean clustering specifically designed for embedded systems. LiDAR sensors are widely used for perception tasks in autonomous vehicles, robotics, and other applications, providing dense point cloud data of the environment. Segmentation of the point cloud into meaningful objects is essential for understanding the surroundings and making informed decisions. Euclidean clustering is an effective approach for grouping points based on spatial proximity, enabling object segmentation. However, implementing such algorithms on embedded systems poses challenges due to limited computational resources. In this work, we propose an optimized implementation of the Euclidean clustering algorithm tailored for embedded systems to achieve real-time performance on the embedded system. The proposed approach involves acquiring raw point cloud data from the LiDAR sensor and preprocessing it to reduce noise and size. Adaptive Euclidean clustering is then applied to group points into clusters based on their spatial proximity. Extracted features such as centroids and bounding boxes are utilized for object classification and segmentation. Post-processing steps refine the segmentation results, improving accuracy and removing spurious clusters

THSI-RP: A two-tier hybrid swarm intelligence based node clustering and multi-hop routing protocol optimization for wireless sensor networks

Wireless sensor networks exploit clustering and routing techniques to improve energy efficiency, but these methods are generally considered as a non-deterministic polynomial (NP-hard) problem. To tackle this problem, we propose a novel two-tier hybrid swarm intelligence-based hierarchical routing protocol (THSI-RP). In the first tier, THIS-RP incorporates a hybrid swarm intelligence algorithm that combines Grey Wolf Optimization (GWO) and Marine Predators Algorithm (MPA) for the clustering algorithm. By incorporating the wolf factors from GWO into MPA as elite predators, we enhance the network search efficiency. This algorithm achieves adaptive optimal clustering with a controllable scale, considering the residual energy of nodes, relative distance between nodes, and node centrality. In the second tier, the routing algorithm uses a hybrid SI algorithm based on GWO and the graph model. First, we integrate distance and energy factors to dynamically select forwarding nodes using GWO. Then, based on the distance and energy balance principle, a weight cost function is established and combined with the minimum spanning tree method to construct a communication routing tree between the forwarding node and the base station to achieve optimized inter-cluster multi-hop routing. Simulation results demonstrate that THSI-RP outperforms several typical routing protocols in various metrics over EEUC, LPSO, LGWO, and LACO. For instance, THSI-RP achieves percentage improvement total packets by 290%, 10.9%, 8.9%, 19.9%, respectively. It also performs better in FND by 88.7%, 10.9%, 8.6%, and 16.3%, in HDN by 86.6%, 8.1%, 6.3%, and 17.1%, and in ADN by 41.7%, 8.2%, 6.4%, 18.3%.

A three-way adaptive density peak clustering (3W-ADPC) method

A three-way adaptive density peak clustering (3W-ADPC) method

Full-parameter adaptive fuzzy clustering for noise image segmentation based on non-local and local spatial information

The fuzzy clustering C-means (FCM) algorithm is a compelling image segmentation method in image segmentation. However, the algorithm is less robust for images containing noise. This paper proposes a non-local full-parameter adaptive spatial information combined with a local fuzzy factor for noisy image segmentation. Firstly, we realize the adaptive calculation of the smoothing parameter, search term window, and neighborhood window in the non-local spatial information by defining the smoothness and designing the adaptive matching function. Secondly, the image’s non-local and local spatial information is considered comprehensively to reduce noise interference and segmentation ambiguity. Finally, the weighted average membership linking is used as the denominator of the objective function to reduce the number of iterations. The results in synthetic noise image and color image segmentation experiments show that the proposed algorithm has outstanding performance in various evaluation metrics and visual effects, outperforming most other variants of fuzzy clustering-based algorithms.

ADSVAE: An Adaptive Density-aware Spectral Clustering Method for Multi-omics Data Based on Variational Autoencoder

Introduction: The discovery of tumor subtypes helps to explore tumor pathogenesis, determine the operability of clinical treatment, and improve patient survival. Clustering analysis is increasingly applied to multi-genomic data. However, due to the diversity and complexity of multi-omics data, developing a complete clustering algorithm for tumor molecular typing is still challenging. Methods: In this study, we present an adaptive density-aware spectral clustering method based on a variational autoencoder (ADSVAE). ADSVAE first learns the underlying spatial information of each omics data using a variational autoencoder (VAE) based on the Wasserstein distance metric. Secondly, a similarity matrix is built for each gene set using an adaptive density-aware kernel. Thirdly, tensor product graphs (TPGs) are used to merge different data sources and reduce noise. Finally, ADSVAE employs a spectral clustering algorithm and utilizes the Gaussian mixture model (GMM) to cluster the final eigenvector matrix to identify cancer subtypes. Results: We tested ADSVAE on 5 TCGA datasets, all with good performance in comparison with several advanced multi-omics clustering algorithms. Compared with the existing multi-group clustering algorithms, the variational autoencoder based on the Wasserstein distance measure in the ADSVAE algorithm can learn the underlying spatial information on each omics data, which has a better effect on learning complex data distribution. The self-tuning density-aware kernel used by the ADSVAE algorithm enhances the similarity between shared near neighbor points, and the process of tensor product plot data integration and diffusion can better reduce the noise and reveal the underlying structure, improving the performance. Conclusion: Due to the inherent pitfalls of computational biology in the study of cancer subtype identification, although some research conclusions have been made in this paper on the related issues, as the research in related fields continues to deepen, the clustering study of cancer subtype identification based on genomic data needs further improvement and refinement.

Fast screening of lithium-ion batteries for second use with pack-level testing and machine learning

Fast and accurate screening of retired lithium-ion batteries is critical to an efficient and reliable second use with improved performance consistency, contributing to the sustainability of renewable energy sources. However, time-consuming testing, representative criteria extraction, and large module-to-module inconsistencies at the end of first life all pose great challenges for fast screening. This paper proposes a fast screening approach with pack-level testing and machine learning to evaluate and classify module-level aging, where disassembly of the battery pack and individual testing of modules are not required. Dynamic characteristic-based criteria are designed to extract the comprehensive performance of the retired modules, making the approach applicable for battery packs with module state-of-charge inconsistencies up to 30%. Adaptive affinity propagation clustering is utilized to classify the modules and further accelerate the screening progress. The proposed approach is implemented and validated by conducting pack-level and module-level experiments with a retired battery pack consisting of 95 modules connected in series. The proposed approach screening time is reduced by at least 50% compared with approaches that require module-level testing. Reasonable static performance consistency and better dynamic performance consistency, as well as higher screening stability, are achieved, achieving the average overall performance improvement of 18.94%, 4.83% and 34.41% compared with the three benchmarks, respectively. Its adaptability to a larger current rate shows promise for large-scale applications in second-use screening.

Multi-Label Classification via Adaptive Resonance Theory-Based Clustering.

This paper proposes a multi-label classification algorithm capable of continual learning by applying an Adaptive Resonance Theory (ART)-based clustering algorithm and the Bayesian approach for label probability computation. The ART-based clustering algorithm adaptively and continually generates prototype nodes corresponding to given data, and the generated nodes are used as classifiers. The label probability computation independently counts the number of label appearances for each class and calculates the Bayesian probabilities. Thus, the label probability computation can cope with an increase in the number of labels. Experimental results with synthetic and real-world multi-label datasets show that the proposed algorithm has competitive classification performance to other well-known algorithms while realizing continual learning.

An aerial targets grouping method based on relation network

AbstractAerial targets grouping is an important content of air battlefield situation awareness, and intelligent aerial targets grouping is an urgent need for intelligent air combat. Traditional aerial targets grouping methods have problems such as difficulty extracting air target features, difficulty designing formation similarity measures, and lack of data samples. To solve the above problems, this paper proposes a relational network based aerial targets grouping method to achieve end‐to‐end grouping under small sample conditions. Firstly, an adaptive clustering algorithm is proposed to preliminarily cluster the aerial targets in the air battlefield situation to obtain the group to be tested. Then, a representation of aerial target groups is designed, which unifies the multi‐dimensional features of the group's temporal, spatial, and attribute. Next, based on this representation, combining 3D convolutional neural networks (3DCNN) and long short‐term memory (LSTM), a neural network structure that can automatically extract multi‐dimensional features of groups is designed. After that, the relation module is used to learn the distance metric between groups. After training the relation network, the type of group to be tested can be identified, thereby achieving the goal of grouping. A dataset of 30 types of aerial target group patterns was established through simulation. The ablation experiments were conducted to verify the performance of each module, and the experimental results showed that the designed feature extraction module can effectively extract temporal sample features, while the designed relationship module performs better in extracting temporal sample information. By comparing several typical traditional aerial target grouping methods, the proposed method improves the grouping accuracy by 11.2% to 13.4%, achieving an accuracy of 91.1% in the 5‐way 1‐shot task and 95.6% in the 5‐way 5‐shot task.

An automated hybrid attention based deep convolutional capsule with weighted autoencoder approach for skin cancer classification

ABSTRACT Skin cancer is a serious cancer caused by the uncontrollable growth of damaged DNA that leads to death. It is essential to identify the disease at the initial stage and eliminate it from spreading. Hence, this research introduces an automated hybrid deep learning (DL) technique for improving the accuracy of cancer diagnostic systems. In the pre-processing, histogram stretching, colour constancy, hair removal and noise elimination process are undertaken. Then, the Adaptive fuzzy c-means clustering (AFC) is introduced for segmenting the tumour portion. Then, the feature extraction and classification stage is performed using an attention-based deep convolutional capsule weighted auto-encoder classifier network (A-DCCN-WAE) technique. For experimentation, the dataset is collected from International Skin Imaging Collaboration (ISIC) 2019 dataset and is implemented in the PYTHON platform. The proposed method obtains an accuracy of 97%, Precision of 95.6%, F-measure of 97.2%, Mathew's correlation coefficient (MCC) of 90.6% and specificity of 96.9%.

A dynamic adaptive iterative clustered federated learning scheme

Clustered federated learning (CFL), as an important research branch of personalized federated learning (FL), can better cope with the highly statistically heterogeneous federated learning environment and provide higher quality services to clients. However, existing CFL schemes have difficulties in adapting to real-time data distribution changes due to disadvantages such as relatively fixed cluster structure. This poses a great challenge to the practical deployment of CFL schemes. To address the common problems of existing CFL schemes, we propose a more flexible dynamic adaptive cluster federated learning scheme (AICFL). AICFL uses the mutual sensitivity between models and data as intuition to perform cluster identity estimation, cluster addition, cluster model updating, and cluster deletion in the early iterations of FL to find the optimal client cluster partitioning. Firstly, this process does not require a priori estimation of the number of clusters and does not require the online participation of all clients. Secondly, during cluster partitioning, AICFL is able to adjust the cluster structure in real time based on the overall data distribution. Moreover, AICFL has the same ability to adapt to changes in the system environment in the middle and late stages of FL. The experimental results show that our scheme gives the most reasonable cluster partitioning results in all cases which indicates that AICFL is able to cope with the above-mentioned distribution changes well, and has better adaptability and better flexibility than other schemes.

Graph structure learning layer and its graph convolution clustering application

To learn the embedding representation of graph structure data corrupted by noise and outliers, existing graph structure learning networks usually follow the two-step paradigm, i.e., constructing a “good” graph structure and achieving the message passing for signals supported on the learned graph. However, the data corrupted by noise may make the learned graph structure unreliable. In this paper, we propose an adaptive graph convolutional clustering network that alternatively adjusts the graph structure and node representation layer-by-layer with back-propagation. Specifically, we design a Graph Structure Learning layer before each Graph Convolutional layer to learn the sparse graph structure from the node representations, where the graph structure is implicitly determined by the solution to the optimal self-expression problem. This is one of the first works that uses an optimization process as a Graph Network layer, which is obviously different from the function operation in traditional deep learning layers. An efficient iterative optimization algorithm is given to solve the optimal self-expression problem in the Graph Structure Learning layer. Experimental results show that the proposed method can effectively defend the negative effects of inaccurate graph structures. The code is available at https://github.com/HeXiax/SSGNN.

Restructuring Graph for Higher Homophily via Adaptive Spectral Clustering

While a growing body of literature has been studying new Graph Neural Networks (GNNs) that work on both homophilic and heterophilic graphs, little has been done on adapting classical GNNs to less-homophilic graphs. Although the ability to handle less-homophilic graphs is restricted, classical GNNs still stand out in several nice properties such as efficiency, simplicity, and explainability. In this work, we propose a novel graph restructuring method that can be integrated into any type of GNNs, including classical GNNs, to leverage the benefits of existing GNNs while alleviating their limitations. Our contribution is threefold: a) learning the weight of pseudo-eigenvectors for an adaptive spectral clustering that aligns well with known node labels, b) proposing a new density-aware homophilic metric that is robust to label imbalance, and c) reconstructing the adjacency matrix based on the result of adaptive spectral clustering to maximize the homophilic scores. The experimental results show that our graph restructuring method can significantly boost the performance of six classical GNNs by an average of 25% on less-homophilic graphs. The boosted performance is comparable to state-of-the-art methods.

ACCD: An Adaptive Clustering-Based Collusion Detector in Crowdsourcing (Student Abstract)

Crowdsourcing is a popular method for crowd workers to collaborate on tasks. However, workers coordinate and share answers during the crowdsourcing process. The term for this is "collusion". Copies from others and repeated submissions are detrimental to the quality of the assignments. The majority of the existing research on collusion detection is limited to ground truth problems (e.g., labeling tasks) and requires a predetermined threshold to be established in advance. In this paper, we aim to detect collusion behavior of workers in an adaptive way, and propose an Adaptive Clustering Based Collusion Detection approach (ACCD) for a broad range of task types and data types solved via crowdsourcing (e.g., continuous rating with or without distributions). Extensive experiments on both real-world and synthetic datasets show the superiority of ACCD over state-of-the-art approaches.