Iterative Clustering Research Articles

Optimizing E-Commerce with Ensemble Learning and Iterative Clustering for Superior Product Selection

Optimizing E-Commerce with Ensemble Learning and Iterative Clustering for Superior Product Selection

Discovering Electric Vehicle Charging Locations Based on Clustering Techniques Applied to Vehicular Mobility Datasets

With the proliferation of vehicular mobility traces because of inexpensive on-board sensors and smartphones, utilizing them to further understand road movements have become easily accessible. These huge numbers of vehicular traces can be utilized to determine where to enhance road infrastructures such as the deployment of electric vehicle (EV) charging stations. As more EVs are plying today’s roads, the driving anxiety is minimized with the presence of sufficient charging stations. By correctly extracting the various transportation parameters from a given dataset, one can design an adequate and adaptive EV charging network that can provide comfort and convenience for the movement of people and goods from one point to another. In this study, we determined the possible EV charging station locations based on an urban city’s vehicular capacity distribution obtained from taxi and ride-hailing mobility GPS traces. To achieve this, we first transformed the dynamic vehicular environment based on vehicular capacity into its equivalent urban single snapshot. We then obtained the various traffic zone distributions by initially utilizing k-means clustering to allow flexibility in the total number of wanted traffic zones in each dataset. In each traffic zone, iterative clustering techniques employing Density-based Spatial Clustering of Applications with Noise (DBSCAN) or clustering by fast search and find of density peaks (CFS) revealed various area separation where EV chargers were needed. Finally, to find the exact location of the EV charging station, we last ran k-means to locate centroids, depending on the constraint on how many EV chargers were needed. Extensive simulations revealed the strengths and weaknesses of the clustering methods when applied to our datasets. We utilized the silhouette and Calinski–Harabasz indices to measure the validity of cluster formations. We also measured the inter-station distances to understand the closeness of the locations of EV chargers. Our study shows how CFS + k-means clustering techniques are able to pinpoint EV charger locations. However, when utilizing DBSCAN initially, the results did not present any notable outcome.

Image segmentation for pest detection of crop leaves by improvement of regional convolutional neural network

Pest detection is important for crop cultivation. Crop leaf is the main place of pest invasion. Current technologies to detect crop pests have constraints, such as low efficiency, storage demands and limited precision. Image segmentation is a fast and efficient computer-aided detection technology. High resolution image capture solidly supports the crucial processes in discerning pests from images. Study of analytical methods help parse information in the images. In this paper, a regional convolutional neural network (R-CNN) architecture is designed in combination with the radial bisymmetric divergence (RBD) method for enhancing the efficiency of image segmentation. As a special application of RBD, the hierarchical mask (HM) is produced to endorse detection and classification of the leaf-dwelling pests, offering enhanced efficiency and reduced storage requirements. Moreover, to deal with some mislabeled data, a threshold variable is introduced to adjust a fault-tolerant mechanism into HM, to generate a novel threshold-based hierarchical mask (TbHM). Consequently, the hierarchical mask R-CNN (HM-R-CNN) model and the threshold-based hierarchical mask R-CNN (TbHM-R-CNN) model are established to detect various types of healthy and pest-invasive crop leaves to select the regional image features that are rich in pest information. Then simple linear iterative clustering (SLIC) method is incorporation to finish the image segmentation for the classification of pest invasion. The models are tuned and optimized, then validated. The most optimized modeling results are from the TbHM-R-CNN model, with the classification accuracy of 96.2%, the recall of 97.5% and the F1 score of 0.982. Additionally, the HM-R-CNN model observed appreciable results second only to the best model. These results indicate that the proposed methodologies are well-suited for training and testing a dataset of plant diseases, offering heightened accuracy in pest classification. This study revealed that the proposed methods significantly outperform the existing techniques, marking a substantial improvement over current methods.

Adaptive superpixel segmentation and pigment identification of colored relics based on visible spectral images

This work explores the extraction of the spatial distribution and chemical composition information of pigments in colored relics through visible spectral images. An adaptive superpixel segmentation method is proposed to extract the spatial distribution information of pigments. Quadtree decomposition is applied to generate nonuniform initial seed points based on the image homogeneity. These seed points are used as the initial cluster centers in an extended simple linear iterative clustering (SLIC) algorithm for visible spectral images to create superpixels of varying sizes that reflect the homogeneity. Each superpixel is subsequently treated as an individual area in the colored relics, and a pigment identification method based on the visible spectral reflectance is proposed to identify the pigments in these areas. A standard reference database is constructed using samples that simulate the painting process of ancient wall paintings in the Mogao Grottoes. Geometric features, which are characterized by the linear combination of the normalized visible spectral reflectance, its slope and its curvature, are designed to represent the chemical composition of pigments. The geometric features of the superpixels are compared with those of the pigments in the database using the Euclidean distance to determine the pigments in each area of the colored relics. This work is expected to provide scientific guidance for pigment selection in the color restoration of colored relics.

Iterative Multiview Subspace Learning for Unpaired Multiview Clustering.

In real applications, several unpredictable or uncertain factors could result in unpaired multiview data, i.e., the observed samples between views cannot be matched. Since joint clustering among views is more effective than individual clustering in each view, we investigate unpaired multiview clustering (UMC), which is a valuable but insufficiently studied problem. Due to lack of matched samples between views, we could fail to build the connection between views. Therefore, we aim to learn the latent subspace shared by views. However, existing multiview subspace learning methods usually rely on the matched samples between views. To address this issue, we propose an iterative multiview subspace learning strategy [iterative unpaired multiview clustering (IUMC)], aiming to learn a complete and consistent subspace representation among views for UMC. Moreover, based on IUMC, we design two effective UMC methods: 1) Iterative unpaired multiview clustering via covariance matrix alignment (IUMC-CA) that further aligns the covariance matrix of subspace representations and then performs clustering on the subspace and 2) iterative unpaired multiview clustering via one-stage clustering assignments (IUMC-CY) that performs one-stage multiview clustering (MVC) by replacing the subspace representations with clustering assignments. Extensive experiments show the excellent performance of our methods for UMC, compared with the state-of-the-art methods. Also, the clustering performance of observed samples in each view can be considerably improved by those observed samples from the other views. In addition, our methods have good applicability in incomplete MVC.

Complete Active Space Iterative Coupled Cluster Theory.

In this work, we investigate the possibility of improving multireference-driven coupled cluster (CC) approaches with an algorithm that iteratively combines complete active space (CAS) calculations with tailored CC and externally corrected CC. This is accomplished by establishing a feedback loop between the CC and CAS parts of a calculation through a similarity transformation of the Hamiltonian with those CC amplitudes that are not encompassed by the active space. We denote this approach as the complete active space iterative coupled cluster (CASiCC) ansatz. We investigate its efficiency and accuracy in the singles and doubles approximation by studying the prototypical molecules H4, H8, H2O, and N2. Our results demonstrate that CASiCC systematically improves on the single-reference CCSD and the externally corrected CCSD methods across entire potential energy curves while retaining modest computational costs. However, the tailored coupled cluster method shows superior performance in the strong correlation regime, suggesting that its accuracy is based on error compensation. We find that the iterative versions of externally corrected and tailored coupled cluster methods converge to the same results.

SoFL: Clustered Federated Learning Based on Dual Clustering for Heterogeneous Data

Federated Learning (FL) is an emerging privacy-preserving technology that enables training a global model beneficial to all participants without sharing their data. However, differences in data distributions among participants may undermine the stability and accuracy of the global model. To address this challenge, recent research proposes client clustering based on data distribution similarity, generating independent models for each cluster in order to enhance FL performance. Nevertheless, due to the uncertainty of participant identities, FL struggles to rapidly and accurately determine the clusters. Most of the existing algorithms distinguish clients by iterative clustering, which not only increases the computing cost of the server but also affects the convergence speed of the federation model. To address these shortcomings, in this paper, we propose a novel clustering-based FL method, SoFL. SoFL introduces SOM networks, improves the quality of cluster data, and eliminates redundant categories through secondary clustering, encouraging more similar clients to train together. Through this mechanism, SoFL completes the clustering task in one round of communication and speeds up the convergence of federated model training. Simulation results demonstrate that SoFL accurately and swiftly adapts to determine the clusters. In different non-IID settings, SoFL’s model accuracy improvements ranged from 9 to 18% compared to FedAvg and FedProx.

Hierarchical homogeneity-based superpixel segmentation: application to hyperspectral image analysis

ABSTRACT Hyperspectral image (HI) analysis approaches have recently become increasingly complex and sophisticated. Recently, the combination of spectral-spatial information and superpixel techniques have addressed some hyperspectral data issues, such as the higher spatial variability of spectral signatures and dimensionality of the data. However, most existing superpixel approaches do not account for specific HI characteristics resulting from its high spectral dimension. In this work, we propose a multiscale superpixel method that is computationally efficient for processing hyperspectral data. The Simple Linear Iterative Clustering (SLIC) oversegmentation algorithm, on which the technique is based, has been extended hierarchically. Using a novel robust homogeneity testing, the proposed hierarchical approach leads to superpixels of variable sizes but with higher spectral homogeneity when compared to the classical SLIC segmentation. For validation, the proposed homogeneity-based hierarchical method was applied as a preprocessing step in the spectral unmixing and classification tasks carried out using, respectively, the Multiscale sparse Unmixing Algorithm (MUA) and the CNN-Enhanced Graph Convolutional Network (CEGCN) methods. Simulation results with both synthetic and real data show that the technique is competitive with state-of-the-art solutions.

Infrared superpixel patch-image model for small target detection under complex background

The main problem of infrared small target detection in complex background is how to effectively eliminate the edge residue. In this paper, we propose an efficient method named Superpixel Patch Image (SPI) model to handle this challenging task. The SPI model can fit the edges of the background well, thus effectively eliminating edge interference in the process of target detection, and achieving excellent performance. The SPI method consists of three steps: Firstly, an improved Simple Linear Iterative Clustering (ISLIC) algorithm is proposed to generate compact superpixels that perfectly match the background edge. Secondly, setting each superpixel patch as a column, a large patch-image matrix is constructed, and the target foreground image and background image is separated by imprecisely augmented Lagrange multiplication. Finally, based on the comprehensively analysis of the distribution characteristics of the target and the highlighted edge in the foreground image, an adaptive threshold is used to extract the target from the foreground superpixel patch. The experimental results of real infrared scenes show that the presented SPI model achieves the best SCRG, BSF and ROC curves compared with the existing 9 state-of-art algorithms, and can effectively extract small targets under different complex backgrounds.

A high-precision automatic extraction method for shedding diseases of painted cultural relics based on three-dimensional fine color model

In recent years, with the development of 3D digitization of cultural relics, most cultural sites contain a large number of fine 3D data of cultural relics, especially complex geometric objects such as painted cultural relics. At present, how to automatically extract surface damage information from the fine 3D color model of painted cultural relics and avoid the loss of accuracy caused by reducing the dimension using conventional methods is an urgentproblem. In view of the above issues, this paper proposes an automatic and high-precision extraction method for cultural relics surface shedding diseases based on 3D fine data. First, this paper designs a 2D and 3D integrated data conversion model based on OpenSceneGraph, a 3D engine, which performs mutual conversion between 3D color model textures and 2D images. Second, this paper proposes a simple linear iterative clustering segmentation algorithm with an adaptive k value, which solves the problem of setting the superpixel k value and improves the accuracy of image segmentation. Finally, through the 2D and 3D integrated models, the disease is statistically analyzed and labeled on the 3D model. Experiments show that for painted plastic objects with complex surfaces, the disease extraction method based on the 3D fine model proposed in this paper has improved geometric accuracy compared with the current popular orthophoto extraction method, and the disease investigation is more comprehensive. Compared with the current 3D manual extraction method in commercial software, this method greatly improves the efficiency of disease extraction while ensuring extraction accuracy. The research method of this paper activates many existing 3D fine data of cultural protection units and converts conventional 2D data mining and analysis into 3D, which is more in line with the scientific utilization of data in terms of accuracy and efficiency and has certain scientific research value, leading value and practical significance.

Efficient clustering of large molecular libraries.

The widespread use of Machine Learning (ML) techniques in chemical applications has come with the pressing need to analyze extremely large molecular libraries. In particular, clustering remains one of the most common tools to dissect the chemical space. Unfortunately, most current approaches present unfavorable time and memory scaling, which makes them unsuitable to handle million- and billion-sized sets. Here, we propose to bypass these problems with a time- and memory-efficient clustering algorithm, BitBIRCH. This method uses a tree structure similar to the one found in the Balanced Iterative Reducing and Clustering using Hierarchies (BIRCH) algorithm to ensure time scaling. BitBIRCH leverages the instant similarity (iSIM) formalism to process binary fingerprints, allowing the use of Tanimoto similarity, and reducing memory requirements. Our tests show that BitBIRCH is already > 1,000 times faster than standard implementations of the Taylor-Butina clustering for libraries with 1,500,000 molecules. BitBIRCH increases efficiency without compromising the quality of the resulting clusters. We explore strategies to handle large sets, which we applied in the clustering of one billion molecules under 5 hours using a parallel/iterative BitBIRCH approximation.

Layer-by-layer unsupervised clustering of statistically relevant fluctuations in noisy time-series data of complex dynamical systems

Complex systems are typically characterized by intricate internal dynamics that are often hard to elucidate. Ideally, this requires methods that allow to detect and classify in an unsupervised way the microscopic dynamical events occurring in the system. However, decoupling statistically relevant fluctuations from the internal noise remains most often nontrivial. Here, we describe "Onion Clustering": a simple, iterative unsupervised clustering method that efficiently detects and classifies statistically relevant fluctuations in noisy time-series data. We demonstrate its efficiency by analyzing simulation and experimental trajectories of various systems with complex internal dynamics, ranging from the atomic- to the microscopic-scale, in- and out-of-equilibrium. The method is based on an iterative detect-classify-archive approach. In a similar way as peeling the external (evident) layer of an onion reveals the internal hidden ones, the method performs a first detection/classification of the most populated dynamical environment in the system and of its characteristic noise. The signal of such dynamical cluster is then removed from the time-series data and the remaining part, cleared-out from its noise, is analyzed again. At every iteration, the detection of hidden dynamical subdomains is facilitated by an increasing (and adaptive) relevance-to-noise ratio. The process iterates until no new dynamical domains can be uncovered, revealing, as an output, the number of clusters that can be effectively distinguished/classified in a statistically robust way as a function of the time-resolution of the analysis. Onion Clustering is general and benefits from clear-cut physical interpretability. We expect that it will help analyzing a variety of complex dynamical systems and time-series data.

Optimization of UWB indoor positioning based on hardware accelerated Fuzzy ISODATA

With the development of wireless communication technology, Ultra-Wideband (UWB) has become an important solution for indoor positioning. In complex indoor environments, the influence of non-line-of-sight (NLOS) factors leads to increased positioning errors. To improve the positioning accuracy, fuzzy iterative self-organizing data analysis clustering algorithm (ISODATA) is introduced to process a large amount of UWB data to reduce the influence of NLOS factors, and to stabilize positioning error within 2 cm, enhances the accuracy of the positioning system. To further improve the running efficiency of the algorithm, FPGA is used to accelerate the key computational part of the algorithm, compared with running on the MATLAB platform, which improves the speed about 100 times, enhances the algorithm’s computational speed dramatically.

A digital twin-based traffic light management system using BIRCH algorithm

Urban transportation networks are vital for the economic and environmental well-being of cities and they are faced with the integration of Human-Driven Vehicles (HVs) and Connected and Autonomous Vehicles (CAVs) challenge. Most of the traditional traffic management systems fail to effectively manage the dynamic and complex flows of mixed traffic, mainly because of large computational requirements and the restrictions that control models of traffic lights directly based on extensive and continuous training data. Most of the times, the operational flexibility of CAVs is severely compromised for the safety of HVs, or CAVs are given high priority without taking into account the efficiency of HVs leading to lower performance, especially at low CAV penetration rates. On the other hand, the existing adaptive traffic light approaches were usually partial and could not adapt to the real-time behaviors of the traffic system. Some systems operate with inflexible temporal control plans that cannot react to variations in traffic flow or use adaptive control strategies that are based on a limited set of static traffic conditions. This paper presents a novel traffic light control approach utilizing the BIRCH (Balanced Iterative Reducing and Clustering using Hierarchies) clustering algorithm combined with digital twins for a more adaptive and efficient system. The BIRCH is effective in processing large datasets because it clusters data points incrementally and dynamically into a small set of representatives. The suggested method does not only enable better simulation and prediction of traffic patterns but also makes possible the real-time adaptive control of traffic signals at signalized intersections. It also improves traffic flow, reduces congestion, and minimizes vehicle idling time by adjusting the green and red light durations dynamically based on both real-time and historical traffic data. This approach is assessed under different traffic intensities, which include low, moderate, and high, while efficiency, fuel consumption, and the number of stops are being compared with the traditional and the existing adaptive traffic management systems.

Structure-preserving image filtering with soft power iteration clustering

Structure-preserving image filtering with soft power iteration clustering

Spatial immunophenotyping using multiplexed imaging of immune follicles in secondary lymphoid tissues.

Secondary lymphoid organs (SLOs), including tonsils (TS), lymph nodes (LN), and Peyer's Patches, exhibit complementary immune functions. However, little is known about the spatial organization of immune cells and extracellular matrix (ECM) in the SLOs. Traditional imaging is limited to a few markers, confining our understanding of the differences between the SLOs. Herein, imaging mass cytometry addressed this gap by simultaneously profiling 25-plex proteins in SLO tissues at subcellular resolution. The antibody panel targeted immune, stromal, chemokine, epigenetic, and functional markers. For robust cell identification, a computational workflow SpatialVizPheno was developed to spatially phenotype 999,970 cells using two approaches, including manual gating and semi-supervised gating, iterative clustering, and annotation. LN exhibited the highest density of B cells while the intestinal tissues contained the highest proportion of regulatory and follicular helper T cells. SpatialVizPheno identified the most prevalent interaction between follicular dendritic cells and stromal cells (SCs), plasmablasts/plasma cells, and the SCs across the lymphoid tissues. Collagen-enriched regions were associated with the spatial orientation of B cell follicles in both TS and LN tissues, but not in intestinal lymphoid tissues. Such spatial differences of immunophenotypes and ECM in different SLO tissues can be used to quantify the relationship between cellular organization and ultimate immune responses.

Monocular Vision Sensor-Based Indoor Road and Stair Detection

Road detection is an essential component of indoor robot navigation. Vision sensors have great advantages in road detection as they can provide rich information in terms of environmental perception. In this paper, a monocular vision sensor-based method for indoor road and stair detection is proposed, which detects feasible areas in indoor environments very fast without paying attention to detailed features of walls or other obstacles. More specifically, for a given indoor road image captured by an on-board vision sensor, the simple linear iterative clustering (SLIC) algorithm-based approach for efficient image segmentation is introduced. Then, according to the DBSCAN algorithm, the generated superpixels are clustered to form large areas of view. The initial road area is obtained through a safe window on the middle bottom of the image. In order to achieve a more accurate road segmentation, the initial image is processed by the binary search, edge detection based on the Canny operator and straight-line detection and location based on the Hough transform, which integrates edge and stair information into road detection. Several experiments are performed to evaluate the performance of the proposed method. The experimental results demonstrate that the proposed method could accurately detect road information and staircase information in images and succeeds in addressing the indoor road-detection problem.

Moving object detection via feature extraction and classification

Abstract Foreground segmentation (FS) plays a fundamental and important role in computer vision, but it remains a challenging task in dynamic backgrounds. The supervised method has achieved good results, but the generalization ability needs to be improved. To address this challenge and improve the performance of FS in dynamic scenarios, a simple yet effective method has been proposed that leverages superpixel features and a one-dimensional convolution neural network (1D-CNN) named SPF-CNN. SPF-CNN involves several steps. First, the coined Iterated Robust CUR (IRCUR) is utilized to obtain candidate foregrounds for an image sequence. Simultaneously, the image sequence is segmented using simple linear iterative clustering. Next, the proposed feature extraction approach is applied to the candidate matrix region corresponding to the superpixel block. Finally, the 1D-CNN is trained using the obtained superpixel features. Experimental results demonstrate the effectiveness of SPF-CNN, which also exhibits strong generalization capabilities. The average F1-score reaches 0.83.

A graph learning approach for automatic subsurface defects segmentation in building façade using RGB and infrared images

Infrared thermal imaging technology provides a non-destructive and automated inspection solution for analyzing building facades. However, conventional methods that rely on detecting subsurface defects through abnormal heat loss often suffer from the limitation of misidentifying other materials present on the building facade. To address this challenge, this study proposes a graph learning approach for segmentation of subsurface defects by integrating both visual and thermal information. The proposed methodology encompasses several key steps. Firstly, the infrared image is partitioned into internally connected superpixels with similar characteristics using a linear iterative clustering algorithm. Subsequently, these superpixels are employed as nodes to construct a weighted undirected graph. The edges of the graph are defined based on the connectivity of superpixels in the image plane, while the weights are determined by the Wasserstein distance between each node. Furthermore, a multimodal graph reuse technique is introduced to construct a new graph based on an RGB image, thereby incorporating additional visual information. By defining the defect as tightly connected in the RGB graph but loosely connected in the infrared graph, the segmentation of defects involving multiple materials is formulated as a multi-graph N-cut problem. The feasibility and effectiveness of the proposed method are demonstrated through laboratory testing conducted on concrete specimens, as well as on-site monitoring carried out on a tiled building façade.

An Identification Method for Anomaly Types of Active Distribution Network Based on Data Mining

With the increasing penetration of distributed generators (DGs) and the growing demand for reliable power sources, it has become imperative to promptly identify anomalies in active distribution networks (ADNs). Additionally, anomaly identification is also crucial to assisting in targeting maintenance to handle failures after the action of relay protections. This study presents a highly accurate and rapidly responsive approach to identifying anomalies in ADNs. The proposed method uses three layers of intricately interconnected modules. In the first module, an iterative clustering layer is devised to process the original data. In the second module, an association rule layer is designed to identify hidden patterns based on the types of anomalies and the post-fault voltage. In the third module, a regression training layer is employed to train accurate identification models. The case study utilizes data samples collected by an IEEE 33-bus system with distributed generations as well as post-fault data from a practical distribution network. The results of comparative tests demonstrate that the proposed data-mining structure achieves reliable performance with high identification accuracy and short reaction times, indicating its ability to be applied to the state awareness system of ADNs.