Clustering Fusion Research Articles

Time-series water prediction based on feature clustering fusion and XGboost

Abstract. The time series prediction problem has a very wide range of applications in many fields. Most scholars use the LSTM class of algorithms for prediction. However, this method consumes a significant amount of computational power and presents several challenges. To address this issue, this paper proposes a time series forecasting method based on feature fusion and XGBoost. Specifically, we first utilize holiday information and the K-Means algorithm for feature extraction to expand the feature dimensions of the dataset, and then employ XGBoost as a model for training and prediction. Experiments demonstrate that the method proposed in this paper significantly reduces error compared to other traditional machine learning and deep learning methods, while the training time is much shorter than these methods. For example, compared with LSTM, the MSLE of this model decreases by 1.42%, while the training time is only 0.15% of that of LSTM. This greatly saves on training costs and computational power consumption. This confirms the effectiveness of using machine learning and clustering algorithms in time series prediction and provides new methods and practical application directions for future time series prediction models.

A Multiview Brain Network Transformer Fusing Individualized Information for Autism Spectrum Disorder Diagnosis.

Functional connectivity (FC) networks, built from analyses of resting-state magnetic resonance imaging (rs-fMRI), serve as efficacious biomarkers for identifying Autism Spectrum Disorders (ASD) patients. Given the neurobiological heterogeneity across individuals and the unique presentation of ASD symptoms, the fusion of individualized information into diagnosis becomes essential. However, this aspect is overlooked in most methods. Furthermore, the existing methods typically focus on studying direct pairwise connections between brain ROIs, while disregarding interactions between indirectly connected neighbors. To overcome above challenges, we build common FC and individualized FC by tangent pearson embedding (TP) and common orthogonal basis extraction (COBE) respectively, and present a novel multiview brain transformer (MBT) aimed at effectively fusing common and indivinformation of subjects. MBT is mainly constructed by transformer layers with diffusion kernel (DK), fusion quality-inspired weighting module (FQW), similarity loss and orthonormal clustering fusion readout module (OCFRead). DK transformer can incorporate higher-order random walk methods to capture wider interactions among indirectly connected brain regions. FQW promotes adaptive fusion of features between views, and similarity loss and OCFRead are placed on the last layer to accomplish the ultimate integration of information. In our method, TP, DK and FQW modules all help to model wider connectivity in the brain that make up for the shortcomings of traditional methods. We conducted experiments on the public ABIDE dataset based on AAL and CC200 respectively. Our framework has shown promising results, outperforming state-of-the-art methods on both templates. This suggests its potential as a valuable approach for clinical ASD diagnosis.

Anti-Rollover Trajectory Planning Method for Heavy Vehicles in Human–Machine Cooperative Driving

The existing trajectory planning research mainly considers the safety of the obstacle avoidance process rather than the anti-rollover requirements of heavy vehicles. When there are driving risks such as rollover and collision, how to coordinate the game relationship between the two is the key technical problem to realizing the anti-rollover trajectory planning under the condition of driving risk triggering. Given the above problems, this paper studies the non-cooperative game model construction method of the obstacle avoidance process that integrates the vehicle driving risk in a complex traffic environment. Then it obtains the obstacle avoidance area that satisfies both the collision and rollover profit requirements based on the Nash equilibrium. A Kmeans-SMOTE risk clustering fusion is proposed in this paper, in which more sampling points are supplemented by the SMOTE oversampling method, and then the ideal obstacle avoidance area is obtained through clustering algorithm fusion to determine the optimal feasible area for obstacle avoidance trajectory planning. On this basis, to solve the convergence problems of the existing multi-objective particle swarm optimization algorithm and analyze the influence of weight parameters and the diversity of the optimization process, this paper proposes an anti-rollover trajectory planning method based on the improved cosine variable weight factor MOPSO algorithm. The simulation results show that the trajectory obtained based on the method proposed in this paper can effectively improve the anti-rollover performance of the controlled vehicle while avoiding obstacles.

DeConFCluster: Deep Convolutional Transform Learning based multiview clustering fusion framework

Multi-view data clustering is essential for discovering patterns and exploiting information from different sources. In this context, we propose DeConFCluster, an unsupervised multi-view clustering fusion framework based on Deep Convolutional Transform Learning (CTL). Our approach has the advantage that it does not require an additional decoder network during the training phase. This makes our model less prone to overfitting in data-constrained scenarios, as opposed to several recent studies based on the encoder–decoder framework. Furthermore, our method incorporates a loss function inspired by K-Means, which enables it to learn more effective representations for the clustering task. Finally, we evaluate our framework on five standard multi-view clustering datasets, and show that it outperforms the state-of-the-art multi-view deep clustering techniques.

Unified Spatial Clustering of Territory Risk to Uncover Impact of COVID-19 Pandemic on Major Coverages of Auto Insurance

This research delves into the fusion of spatial clustering and predictive modeling within auto insurance data analytics. The primary focus of this research is on addressing challenges stemming from the dynamic nature of spatial patterns in multiple accident year claim data, by using spatially constrained clustering. The spatially constrained clustering is implemented under hierarchical clustering with a soft contiguity constraint. It is highly desirable for insurance companies and insurance regulators to be able to make meaningful comparisons of loss patterns obtained from multiple reporting years that summarize multiple accident year loss metrics. By integrating spatial clustering techniques, the study not only improves the credibility of predictive models but also introduces a strategic dimension reduction method that concurrently enhances the interpretability of predictive models used. The evolving nature of spatial patterns over time poses a significant barrier to a better understanding of complex insurance systems as these patterns transform due to various factors. While spatial clustering effectively identifies regions with similar loss data characteristics, maintaining up-to-date clusters is an ongoing challenge. This research underscores the importance of studying spatial patterns of auto insurance claim data across major insurance coverage types, including Accident Benefits (AB), Collision (CL), and Third-Party Liability (TPL). The research offers regulators valuable insights into distinct risk profiles associated with different coverage categories and territories. By leveraging spatial loss data from pre-pandemic and pandemic periods, this study also aims to uncover the impact of the COVID-19 pandemic on auto insurance claims of major coverage types. From this perspective, we observe a statistically significant increase in insurance premiums for CL coverage after the pandemic. The proposed unified spatial clustering method incorporates a relabeling strategy to standardize comparisons across different accident years, contributing to a more robust understanding of the pandemic effects on auto insurance claims. This innovative approach has the potential to significantly influence data visualization and pattern recognition, thereby improving the reliability and interpretability of clustering methods.

Predictive Analytics on Shopee for Optimizing Product Demand Prediction through K-Means Clustering and KNN Algorithm Fusion

Predictive Analytics on Shopee for Optimizing Product Demand Prediction through K-Means Clustering and KNN Algorithm Fusion

Contour Extraction of UAV Point Cloud Based on Neighborhood Geometric Features of Multi-Level Growth Plane

The extraction of UAV building point cloud contour points is the basis for the expression of a three-dimensional lightweight building outline. Previous unmanned aerial vehicle (UAV) building point cloud contour extraction methods have mainly focused on the expression of the roof contour, but did not extract the wall contour. In view of this, an algorithm based on the geometric features of the neighborhood points of the region-growing clustering fusion surface is proposed to extract the boundary points of the UAV building point cloud. Firstly, the region growth plane is fused to obtain a more accurate segmentation plane. Then, the neighboring points are projected onto the neighborhood plane and a vector between the object point and neighborhood point is constructed. Finally, the azimuth of each vector is calculated, and the boundary points of each segmented plane are extracted according to the difference in adjacent azimuths. Experiment results show that the best boundary points can be extracted when the number of adjacent points is 24 and the difference in adjacent azimuths is 120. The proposed method is superior to other methods in the contour extraction of UAV buildings point clouds. Moreover, it can extract not only the building roof contour points, but also the wall contour points, including the window contour points.

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 energy efficient fusing data gathering protocol in wireless sensor networks

The Internet of Things (IoT) is rapidly developing as a promising technology for today’s digital world. Wireless sensor networks (WSN) are a crucial component of IoT, and one of the major problems faced is energy constraints. The clustering protocol is an effective energy saving solution. In this paper, we propose an energy efficient fusing data gathering protocol via distributed second level clustering and deep learning-based data fusion (EEFDG) for WSN. EEFDG divides the WSN into several second-level clusters by a novel distributed cluster algorithm. The transmission distance and the number of neighbors are jointly considered to select an energy-efficient parent node in each second-level cluster. A new cluster head weight function is proposed for intra-cluster parent node rotation. To minimize redundant data transmission, a multivariate convolutional neural network and long short-term memory network (MVCLNet) are designed for data fusion. MVCLNet is deployed hierarchically into the second-level clustered structure to extract features of the multi-sensor data. The EEFDG is an effective solution to the energy limitation problem of WSN. The proposed second-level clustering and data fusion algorithm based on MVCLNet typically decreases communication distances and the amount of data, thereby minimizing communication overhead and energy consumption. Simulation results show that EEFDG significantly prolongs the network lifetime in terms of the first node dead, compared to existing protocols including MR-LEACH, BPDA, CNNDA, FDEAM, DCNN, and LPLL-LEACH, achieving improvements of 181.3%, 42.2%, 48.2%, 19.6%, 62.0%, and 30.2%, respectively. The EEFDG enables highly accurate data fusion and has superior performance in terms of remaining energy, network lifetime, and data collection.

An efficient Analysis of the Fusion of Statistical-Centred Clustering and Machine Learning for WSN Energy Efficiency

Recently, wireless sensor networks on several challenging topics have piqued researchers’ attention. Maximising a network's lifetime requires just the right combination of cluster size and number of nodes. Data transmission from nodes to cluster leaders is energy intensive, even for a modest number of clusters. If there are several clusters, many leaders will be chosen, and many nodes will rely on long-distance transmission to communicate with the home base. Therefore, in order to maximise efficiency, it is necessary to strike a balance between these two factors. WSN's major challenge is improving its energy efficiency. This is because their energy consumption defines their lifespan, and it is difficult, if not impossible, to recharge their batteries. Therefore, it is crucial to develop algorithms that consume as little energy as possible in order to maximise the network's potential. The perfect clusters are essential for the longevity of the network. Therefore, an algorithm called statistical centre energy efficient clustering approach (SEECA) is presented to increase the network's lifetime while decreasing its energy consumption. The experimental findings show that the proposed methodology SCEECA outperforms the LEACH method by a wide margin, with gains of 32% in Residual energy, 16% in Network Lifetime, and 12% in Throughput.

Ocean Wind Speed Prediction Based on the Fusion of Spatial Clustering and an Improved Residual Graph Attention Network

Accurately predicting wind speed is crucial for the generation efficiency of offshore wind energy. This paper proposes an ultra-short-term wind speed prediction method using a graph neural network with a multi-head attention mechanism. The methodology aims to effectively explore the spatio-temporal correlations present in offshore wind speed data to enhance the accuracy of wind speed predictions. Initially, the offshore buoys are organized into a graphical network. Subsequently, in order to cluster the nodes with comparable spatio-temporal features, it clusters the nearby nodes around the target node. Then, a multi-head attention mechanism is incorporated to prioritize the interconnections among distinct regions. In the construction of the graph neural network, a star topology structure is formed by connecting additional nodes to the target node at the center. The effectiveness of this methodology is validated and compared to other time series-based approaches through comparative testing. Metrics such as Mean Absolute Error, Mean Squared Error, Root Mean Squared Error, and R yielded values of 0.364, 0.239, 0.489, and 0.985, respectively. The empirical findings indicate that graph neural networks utilizing a multi-head attention mechanism exhibit notable benefits in the prediction of offshore wind speed, particularly when confronted with intricate marine meteorological circumstances.

Multiple damage segmentation and extraction of wind turbine blades surface under complex background based on SKRT approach

Machine vision detection technology has been widely used in detecting wind turbine blade surface damage, but the complex background often has a significant impact on blade damage detection. Aiming at the complex background blade images caused by various natural background features, complex blade surface features, low contrast features, non-uniform illumination, weather conditions, and other factors, the work proposes an SKRT segmentation method based on K-means clustering and adaptive threshold fusion to segment and extract multiple damage features on the surface of wind turbine blades with complex backgrounds. Firstly, an single-scale retinex(SSR) algorithm is adopted to enhance the contrast between surface damage and image background in the research. Then the image is roughly segmented by K-means clustering to remove most of the background features, and the background mask is replaced with the peak of the remaining feature image pixels to solve the problem of wrong segmentation. Finally, making use of the adaptive threshold local segmentation method to accurately segment the blade surface damage. The experimental results show that the proposed SKRT segmentation method can significantly improve the segmentation accuracy of wind turbine blade surface damage with complex background, and the accuracy, intersection over union, and F-measure value are increased by 37.20%, 35.71%, and 28.69%, respectively. The method performs better robustness to multiple damage feature segmentation.

Fully automated operational modal identification of regular and irregular buildings with ensemble learning

Modal identification, a type of system identification that employs modal analysis, can be automated to reduce user interference in extracting modal parameters. Modal analysis is a valuable tool for obtaining a structure's modal parameters, which define its dynamic characteristics. Operational modal identification is a specific type of modal analysis that uses only system outputs to extract modal parameters. The present study aims to introduce a fully automated modal identification algorithm to identify modal parameters for an unknown system based on its output information. The algorithm uses a stochastic subspace algorithm to extract modal parameters, requiring only the appropriate degree of the state-space model estimated using the Singular Value Criterion. Several features are used to create a feature space that separates physical and spurious modes. The Improved Distance Evaluation algorithm selects appropriate features, making the identification process more accurate and flexible. The identification algorithm employs clustering algorithms with different approaches and the Dempster-Shafer data fusion algorithm to integrate clustering algorithm results and avoid divergence among them. Due to limited data diversity for identification in cases where only one dataset is available, statistical resampling methods are used as an alternative. The proposed algorithm can identify modal parameters using only one dataset of the building structure. Signal processing, data collection, clustering, feature extraction, model reduction, and data fusion are all aspects of this algorithm. To assess its performance, four regular and irregular buildings are considered as case studies. The presented modal identification algorithm is highly accurate and robust with the repeatable result as compared to similar algorithms, with extracted modal parameters demonstrating high stability.

Clustering fusion algorithm for selection of historical beam steady-state transmission data in CAFe facility

Accelerator-driven systems (ADS) are promising technologies for nuclear waste transmutation and energy production. The China ADS Front-end Superconducting Demo Linac (CAFe) is a prototype of the China Initiative Accelerator Driven System (CiADS), which aims to verify the feasibility of key technologies of CiADS. In this article, a novel method for historical data screening of the beam transport in the medium energy beam transport (MEBT) section of CAFe is presented. A clustering fusion algorithm based on unsupervised learning and beam transmission characteristics is designed to extract a large number of samples with the beam in steady-state transmission from historical beam data. A deep neural network model was constructed to fit the beam transport characteristics and verify the reliability of the screened data samples. The method can improve the efficiency and accuracy of data analysis and provide valuable insights for the optimization and control of beam transport in CAFe.

Harnessing Deep Learning Techniques for Text Clustering and Document Categorization

This research paper delves into the realm of deep text clustering algorithms with the aim of enhancing the accuracy of document classification. In recent years, the fusion of deep learning techniques and text clustering has shown promise in extracting meaningful patterns and representations from textual data. This paper provides an in-depth exploration of various deep text clustering methodologies, assessing their efficacy in improving document classification accuracy. Delving into the core of deep text clustering, the paper investigates various feature representation techniques, ranging from conventional word embeddings to contextual embeddings furnished by BERT and GPT models.By critically reviewing and comparing these algorithms, we shed light on their strengths, limitations, and potential applications. Through this comprehensive study, we offer insights into the evolving landscape of document analysis and classification, driven by the power of deep text clustering algorithms.Through an original synthesis of existing literature, this research serves as a beacon for researchers and practitioners in harnessing the prowess of deep learning to enhance the accuracy of document classification endeavors.

Multi-omics approaches identify novel prognostic biomarkers of autophagy in uveal melanoma.

Uveal melanoma (UVM) is a rare yet malignant ocular tumor that metastases in approximately half of all patients, with the majority of those developing metastasis typically succumbing to the disease within a year. Hitherto, no effective treatment for UVM has been identified. Autophagy is a cellular mechanism that has been suggested as an emerging regulatory process for cancer-targeted therapy. Thus, identifying novel prognostic biomarkers of autophagy may help improve future treatment. Consensus clustering and similarity network fusion approaches were performed for classifying UVM patient subgroups. Weighted correlation network analysis was performed for gene module screening and network construction. Gene set variation analysis was used to evaluate the autophagy activity of the UVM subgroups. Kaplan-Meier survival curves (Log-rank test) were performed to analyze patient prognosis. Gene set cancer analysis was used to estimate the level of immune cell infiltration. In this study, we employed multi-omics approaches to classify UVM patient subgroups by molecular and clinical characteristics, ultimately identifying HTR2B, EEF1A2, FEZ1, GRID1, HAP1, and SPHK1 as potential prognostic biomarkers of autophagy in UVM. High expression levels of these markers were associated with poorer patient prognosis and led to reshaping the tumor microenvironment (TME) that promotes tumor progression. We identified six novel potential prognostic biomarkers in UVM, all of which are associated with autophagy and TME. These findings will shed new light on UVM therapy with inhibitors targeting these biomarkers expected to regulate autophagy and reshape the TME, significantly improving UVM treatment outcomes.

Method to Profiling the Characteristics of Indonesian Dangdut Songs, Using K-Means Clustering and Features Fusion

Method to Profiling the Characteristics of Indonesian Dangdut Songs, Using K-Means Clustering and Features Fusion

TWC-EL: A multivariate prediction model by the fusion of three-way clustering and ensemble learning

Multivariate data analysis, as an important research topic in the field of machine learning, focuses on how to utilize the intrinsic connection between feature variables and target variables. However, in the face of complex multivariate prediction environments, existing single prediction models often fail to obtain ideal results. Meanwhile, existing ensemble prediction models are not always adapted to certain complex data. Moreover, the randomness in the clustering process cannot guarantee the clustering accuracy. Therefore, to improve the model’s prediction accuracy and ability to adapt to complex data and reduce the impact of randomness on clustering accuracy, this paper designs a multivariate prediction model utilizing three-way clustering (TWC) and ensemble learning, which is named the TWC-EL model. First, the initial division of the sample set is realized by k-means clustering algorithm, and further the sample set is divided again via the k-means clustering algorithm to solve the problem of clustering accuracy. Then, the results of clustering twice are combined according to the difference in the number of intersection points and the distance from the samples to the center point of each cluster, and the core and fringe regions of each cluster in the initial clustering results are obtained, forming a new TWC method. Next, based on the correlation between the regions, the obtained core and fringe regions are classified into low-correlation, medium-correlation and high-correlation regions, and an ensemble prediction model is designed by combining the advantages of the Elman neural network model, the Extreme Learning Machine (ELM) model and the back propagation neural network (BPNN) model. Finally, the experimental analysis results exhibit that the constructed TWC-EL model is efficient and feasible, and points out the excellent performance compared with the existing prediction models. The validity of the TWC method and the ensemble prediction model in the proposed TWC-EL model are verified by experiments, respectively.

Fusion of Centroid-Based Clustering With Graph Clustering: An Expectation-Maximization-Based Hybrid Clustering.

This article extends the expectation-maximization (EM) formulation for the Gaussian mixture model (GMM) with a novel weighted dissimilarity loss. This extension results in the fusion of two different clustering methods, namely, centroid-based clustering and graph clustering in the same framework in order to leverage their advantages. The fusion of centroid-based clustering and graph clustering results in a simple "soft" asynchronous hybrid clustering method. The proposed algorithm may start as a pure centroid-based clustering algorithm (e.g., k -means), and as the time evolves, it may eventually and gradually turn into a pure graph clustering algorithm [e.g., basic greedy asynchronous distributed interference avoidance (GADIA) (Babadi and Tarokh, 2010)] as the algorithm converges and vice versa. The "hard" version of the proposed hybrid algorithm includes the standard Hopfield neural networks (and, thus, Bruck's Ln algorithm by (Bruck, 1990) and the Ising model in statistical mechanics), Babadi and Tarokh's basic GADIA in 2010, and the standard k -means (Steinhaus, 1956), (MacQueen, 1967) [i.e., the Lloyd algorithm (Lloyd, 1957, 1982)] as its special cases. We call the "hard version" of the proposed clustering as "hybrid-nongreedy asynchronous clustering (H-NAC)." We apply the H-NAC to various clustering problems using well-known benchmark datasets. The computer simulations confirm the superior performance of the H-NAC compared to the k -means clustering, k -GADIA, spectral clustering, and a very recent clustering algorithm structured graph learning (SGL) by Kang et al. (2021), which represents one of the state-of-the-art clustering algorithms.

Effective & near real-time track-to-track association for large sensor data in Maritime Tactical Data System

The Maritime Tactical Data System is a software system that collects track data from maritime sensors to compile and show on a sea map to provide maritime patrol vessels or maritime surveillance stations with situational awareness. Speed and precision in tracking multiple targets are crucial in achieving situational awareness. A multi-target tracking problem is NP-hard if it involves more than two sensors and a large amount of data since it generates many potential solutions that must be evaluated. Previous research has demonstrated that the Density-based Spatial Clustering of Applications with Noise (DBSCAN) method can perform track-to-track association with pretty good results; nevertheless, the density-reachable concept of DBSCAN poses a problem when two targets are within a distance less than the threshold. Another limitation is the inability of DBSCAN to associate tracks as soon as sensor track data is received. DBSCAN must run after all data has been collected in a database. In this paper, a novel track-to-track association method called Neighborhood Clustering Track Association and Fusion (NCTAF) is proposed to address the limitations of DBSCAN. According to the experiment results, NCTAF overcame the inaccurate cluster form generated by DBSCAN. The most remarkable result is that NCTAF performs track associations in an average of one second after receiving sensor track data involving three sensors, 4000 track data per sensor, and an update rate of 5-12 s per sensor. In contrast, DBSCAN required more than 10 min for the same scenario.