Cluster Density Research Articles

Adapting Gaussian Mixture Model Training to Embedded/Edge Devices: A Low I/O, Deadline-Aware and Energy Efficient Design

This work focuses on the Gaussian Mixture Model (GMM), a machine learning model used for density estimation and cluster analysis in healthcare, networking, etc. The Expectation-Maximization (EM) algorithm is commonly used to train GMMs. One of the main challenges facing this algorithm when running on embedded systems is the crippling memory constraints. In fact, EM requires several scans of the dataset and we observed that when the dataset cannot fully reside in the main memory, its execution is dramatically slowed down by I/O movements. In this paper, we present an optimization of the EM algorithm for GMMs that reduces the number of I/O operations thanks to two main contributions: (1) A divide-and-conquer strategy that divides the dataset into chunks, learns the GMM separately on each chunk and combines the results incrementally. By doing so, we prevent data from being swapped several times during the learning process. (2) Restricting the training on a subset of data whose volume is inferred online using data properties while producing good accuracy. On average, our results show a 63% improvement in overall execution time with comparable accuracy. We also adapted GMM learning to run in a limited time budget while hitting a good trade-off between execution time and energy consumption. This solution succeeded in meeting the fixed deadline in 100% of the cases and in reducing the energy consumption by up to 68.77%.

Leader Bagging Improves Lateral Branching and Cropping Potential of ‘Gala’ and ‘WA 38’ Apple during Orchard Establishment

Inadequate lateral branch development can lead to decreased apple (Malus ×domestica Borkh.) orchard productivity and profitability in modern high-density orchard systems. Although plant growth regulator applications are used to increase lateral branching on leaders of young apple trees, inconsistent responses have been observed in the southeastern United States. In North Carolina and Washington, three experiments were conducted to identify effective leader management strategies to increase lateral branching. Effects and interactions of leader bagging, 6-benzyladenine (6-BA), and 6-BA + gibberellic acid (GA4+7) on lateral branch development of 1-year-old leaders were evaluated. Across all experiments, leader bagging was an influential factor. When compared with unbagged trees, leader bagging increased lateral branch number (20% to 48%), number of feathers (74% to 125%), average branch length (28% to 34%), and total linear bearing surface (428%) of the treated section of the leader. Blossom cluster density and final fruit set were increased in bagged trees, 65% and 36%, respectively. At the rates and timings tested, 6-BA and 6-BA + GA4+7 were generally ineffective in stimulating lateral branching and interactions among the factors evaluated were not influential. Leader bagging was an effective lateral branch induction strategy, although the mechanism of action is poorly understood. Future research to characterize the bagged environment and/or physiological responses to bagging may aid in the development of future environmentally sustainable technologies to stimulate lateral branching of apple trees.

STP-KDE: A spatiotemporal trajectory protection and publishing method based on kernel density estimation

Promoted by the Internet of Thing era, the widespread use of mobile sensing devices equipped with positioning functions has led to the generation of substantial trajectory data. Mining and analyzing trajectory data has high research value, but poses a risk of user privacy leakage, resulting in fewer publicly available trajectory datasets for research and analysis. Therefore, a trajectory data publishing method that ensures high data utility while protecting user privacy has become a hot topic. In this paper, we propose a spatiotemporal trajectory data protection and publishing method based on kernel density estimation (STP-KDE), which protects the trajectory data and trajectory count values while improving data utility. In the protection process of trajectory data, we designed a kernel density clustering framework that is combined with the differential privacy exponential mechanism. In the protection process of trajectory counts value, the adaptive Laplace noise perturbation mechanism is proposed to differentially protect these counts. Experimental results show that STP-KDE can provide more useful data and stronger privacy protection than existing studies.

Early-stage silver growth during sputter deposition on SiO2 and polystyrene – Comparison of biased DC magnetron sputtering, high-power impulse magnetron sputtering (HiPIMS) and bipolar HiPIMS

The integration of silver thin films into optoelectronic devices has gained much interest due to their exceptional properties in terms of conductivity and compatibility with flexible substrates. For this type of application, ultra-thin layers are desirable, because of their optical transparency. Standard direct current magnetron sputtering (DCMS) is known to lead to undesirable formation of islands at low effective film thicknesses on typical substrates like SiO2 or polystyrene (PS). Therefore, in this study, we explore high-power impulse magnetron sputtering (HiPIMS) with optional further acceleration of metal ions by biasing the substrate or an additional positive pulse (bipolar HiPIMS) for the fabrication of ultra-thin silver layers. The morphology and electrical properties of ultra-thin silver layers with selected effective thicknesses are characterized on SiO2 and PS substrates. The growth evolution of characteristic parameters is further investigated by in-situ grazing-incidence small-angle X-ray scattering (GISAXS). The results show that HiPIMS deposition yields films with a higher density of clusters than DCMS leading to a percolation threshold at lower effective film thicknesses. This effect is amplified by further ion acceleration. Thus, we suggest HiPIMS as a promising technique for fabricating ultra-thin, conductive layers on organic and oxide substrates.

Connection density based clustering: A graph-based density clustering method

In this paper, a graph-based density clustering framework is proposed that detects the boundary points of clusters rather than cluster exemplars in high density regions. The framework introduces the connection density to measure the density relationship between points, which depends on both a density metric and a distance metric, and a weighted graph is constructed based on the connection density. By cutting off edges with low connection density, points located at the boundary region become isolated points from the rest of the weighted graph, and each connected subgraph forms an initial cluster. For the generated isolated points, a graph-based label propagation strategy is designed, the stable point in initial clusters with higher connection density with an isolated point preferentially propagate labels. Finally, a novel connection density based clustering algorithm is proposed, called CDBC, which can automatically identify clusters of arbitrary shapes. The experimental results show that the proposed method outperforms other advanced clustering algorithms on several synthetic and real-world datasets.

Using Density and Fuzzy Clustering for Data Cleaning and Segmental Description of Livestock Data

Using Density and Fuzzy Clustering for Data Cleaning and Segmental Description of Livestock Data

Linkage among the neutron-skin thickness of 208Pb and nuclear symmetry energy using heavy particle radioactivity

The nuclear symmetry energy (NSE) is a linchpin in deciphering the behavior of matter in a wider domain extending from the characteristics of exotic nuclei to those of neutron stars in the cosmos. Therefore, it is crucial to utilize potential probes to constrain the NSE and its slope parameter L(ρ 0). In this work, we put forth the heavy particle radioactivity (HPR) as a probable bridge among the slope of NSE (L(ρ 0)) and neutron-skin thickness of 208Pb (R skin 208), which serves to put constrain on the L(ρ 0) value. The NSE and its slope parameter are determined from the single nucleon potential of asymmetric nuclear matter exploiting the analytical relationship between these quantities. The isovector/symmetry potential component of the single nucleon potential is derived through HPR for varying R skin 208 by employing the heavy particle/cluster densities and core densities from the relativistic mean field model in conjunction with M3Y nucleon–nucleon interaction. It facilitates in constraining the L(ρ 0) value and neutron skin of finite nuclei using HPR as a linkage, where heavy cluster and core densities of standard Fermi form are considered. The constrained value of L(ρ o ) is 45 ± 8 MeV, which aligns with other estimations derived from nuclear mass measurements, dipole polarizability measurements, and astrophysical data.

Conditional heavy hitter monitoring and application of heterogeneous graph streams based on sketches

Graph streams offer valuable insights into real network evolution in the era of big data. Heavy hitters and conditional heavy hitters focus on identifying entities with significant network traffic, revealing network data characteristics. In this paper, we present two methods, BLSketch and MHLSketch, for identifying heavy hitters and conditional heavy hitters in heterogeneous graph streams. BLSketch uses the bidirectional mapper, Bimap, for efficient query response times while retaining candidate information. MHLSketch uses min heap to preserve a minimal set of candidates to minimize storage space, excelling in space efficiency. The two methods can be incorporated with any sketch based graph summarization method, showing excellent adaptability. Experimental results on four real-world heterogeneous datasets of varying scales with the largest dataset containing millions of edges demonstrate our methods significantly improve the quality of query answering, outperforming state-of-the-art methods, with precision close to 1 and average relative error close to 0. We further combine our methods with density clustering to introduce the SDC model, which efficiently identifies hot regions and key users in Location-Based Social Networks (LBSN). In experiments, SDC outperforms traditional clustering by reducing up to 93.1% of cold Points of Interest (POIs) and shows strong potential for practical applications.

Clustering of passive tracers in a random acoustic velocity field

We consider the effects of passive tracer clustering (e.g., the magnetic field energy in stellar atmospheres) in a random acoustic velocity field. A method for numerical modeling of a two-dimensional random acoustic field is proposed. The field is described by a correlation tensor defined by traveling isotropic waves, taking into account dissipation. Two metrics for measuring the clustering effects are used: concentration and density. Using numerical modeling, we show that the tracer concentration is almost always clustered. The situation with the density is different; as the dissipation tends to zero, the time to reach the clustered states increases significantly. In addition, due to the tracer transport out of the density clustering regions, only a part of the tracer is clustered. For the presented analyses, we considered ensembles of Lagrangian particles and introduced and applied the statistical topography methodology.

Membrane Fluctuation Model for Understanding the Effect of Receptor Nanoclustering on the Activation of Natural Killer Cells through Biomechanical Feedback.

We investigated the role of ligand clustering and density in the activation of natural killer (NK) cells. To that end, we designed reductionist arrays of nanopatterned ligands arranged with different cluster geometries and densities and probed their effects on NK cell activation. We used these arrays as an artificial microenvironment for the stimulation of NK cells and studied the effect of the array geometry on the NK cell immune response. We found that ligand density significantly regulated NK cell activation while ligand clustering had an impact only at a specific density threshold. We also rationalized these findings by introducing a theoretical membrane fluctuation model that considers biomechanical feedback between ligand-receptor bonds and the cell membrane. These findings provide important insight into NK cell mechanobiology, which is fundamentally important and essential for designing immunotherapeutic strategies targeting cancer.

Spatial Analysis of Internally Displaced Persons’ Camps in Borno State, Nigeria

The situation in Borno State, Nigeria has been exacerbated by the ongoing insurgency led by Boko Haram in northeast Nigeria, leading to widespread displacement and the establishment of Internally Displaced Persons (IDP) camps. This study aims to identify the locations of IDP camps in Borno State and analyze their spatial distribution, density, and clustering patterns. Furthermore, it seeks to investigate the geospatial implications of these patterns. To achieve this, the study employs geospatial analysis techniques such as Kernel Density Estimation, Cluster and Outlier Analysis, Nearest Neighbour Analysis, and Spatial Autocorrelation Analysis. The data collected for this study includes location coordinates, satellite imagery, demographic information, and administrative maps. The findings reveal that the distribution of IDP camps in Borno State is non-random and demonstrates varying camp densities across senatorial districts, with predominant high-high clusters and a deliberate, non-random arrangement. These results align with spatial analysis theories, emphasizing the importance of recognizing hidden inequalities in IDP camps and informing targeted interventions and resource planning for a more effective and equitable humanitarian response in Borno State. The implications extend to policy planning and enhanced vulnerability assessment for effective humanitarian responses, stressing the need for tailored interventions based on identified hotspots that address root causes and implement both immediate and long-term solutions. Future research directions could involve more in-depth investigations into the influencing factors, socio-economic impacts, and temporal dynamics of spatial patterns in displacement.

Global aerosol-type classification using a new hybrid algorithm and Aerosol Robotic Network data

Abstract. The properties of aerosols are highly uncertain owing to the complex changes in their composition in different regions. The radiative properties of different aerosol types differ considerably and are vital for studying aerosol regional and/or global climate effects. Traditional aerosol-type identification algorithms, generally based on cluster or empirical analysis methods, are often inaccurate and time-consuming. In response, our study aimed to develop a new aerosol-type classification model using an innovative hybrid algorithm to improve the precision and efficiency of aerosol-type identification. This novel algorithm incorporates an optical database, constructed using the Mie scattering model, and employs a random forest algorithm to classify different aerosol types based on the optical data from the database. The complex refractive index was used as a baseline to assess the performance of our hybrid algorithm against the traditional Gaussian kernel density clustering method for aerosol-type identification. The hybrid algorithm demonstrated impressive consistency rates of 90 %, 85 %, 84 %, 84 %, and 100 % for dust, mixed-coarse (mixed, course-mode aerosol), mixed-fine (mixed, fine-mode aerosol), urban/industrial, and biomass burning aerosols, respectively. Moreover, it achieved remarkable precision, with evaluation metric indexes for micro-precision, micro-recall, micro-F1-score, and accuracy of 95 %, 89 %, 91 %, and 89 %, respectively. Lastly, a global map of aerosol types was generated using the new hybrid algorithm to characterize aerosol types across the five continents. This study, utilizing a novel approach for the classification of aerosol, will help improve the accuracy of aerosol inversion and determine the sources of aerosol pollution.

Graph deep learning recognition of port ship behavior patterns from a network approach

The detection and recognition of ship mobile behavior patterns based on automatic identification system (AIS) data are crucial for ship navigation and maritime supervision. However, present methods, such as rule-based constraints and density clustering extraction, still suffer from restrictions imposed by artificial settings of parameters and thresholds. The lack of neighborhood context information in unstructured and serialized ship trajectory vector data makes it challenging to incorporate advanced methods such as deep learning for pattern learning. This study builds multiple semi-supervised deep learning graph neural network (GNN) models based on the message passing paradigm for recognizing ship mobile behavior patterns by constructing an AIS trajectory network and computing multi-order node features through Delaunay triangulation. It can get rid of the dependence on parameters of traditional rule-based and unsupervised learning methods. It also reduces the dependence on auxiliary information in trajectory qualitative by introducing multi-order node features. We conducted the experiments on cargo ship data from New York and Singapore to test our method. The model's classification accuracy is more than 99% for the New York dataset and more than 95% for the Singapore dataset. This method can be extended to other types of ships and traffic trajectory vector data with the excellent application potential of GNN in pattern recognition of vector mobile data. The code and experimental data are available at: https://github.com/destiny1103/DT-GNN.

The temperature effect on the formation and growth of helium-vacancy clusters in bcc iron: A comprehensive molecular dynamics simulation

Bubble generation can significantly degrade the performance of fusion reactor structural materials. In this work, the temperature effects on the initial clustering of helium (He) atoms, the formation of helium-vacancy clusters (HenVm) and the subsequent growth of the clusters into bubbles in bcc iron under different conditions were simulated by molecular dynamics. The results show that He behaviors are primarily dominated by coalescence at temperature range from 400 K to 500 K, which leads to a preference for He to bind together. Conversely, when temperature is higher than 500 K, dissociation become more prominent, causing large He clusters to be hard to form. Under the conditions of increasing He concentration and irradiation dose, the higher the temperature, the higher the density of HenVm clusters at He concentration around 100 appm. However, the temperature dependence reversed at high He concentration. The detailed analysis of formation cases revealed the percentage of nucleation sites for He bubbles at defects during He segregation. In addition, there is no significant difference among the He/V ratio of HenVm clusters formed at vacancies, interstitial dislocation loops (IDLs) and interstitial sites during the cluster growth. However, the He/V ratio of He bubbles growing on edge dislocation (ED) is overall lower than those growing at other formation sites, since the growth of HenVm clusters on ED requires a lower pressure to kick out self-interstitial atoms (SIAs) or SIA clusters. The average He/V ratio of He bubbles in bulk decreases as temperature rises from 400 K to 800 K, while that of He bubbles growing on ED is independent of temperature.

Low-Temperature and High-Efficiency Solid-Phase Amplification Based on Formamide.

The thermal stability of DNA immobilized on a solid surface is one of the factors that affects the efficiency of solid-phase amplification (SP-PCR). Although variable temperature amplification ensures high specificity of the reaction by precisely controlling temperature changes, excessively high temperatures during denaturation can negatively affect DNA stability. Formamide (FA) enables DNA denaturation at lower temperatures, showing potential for SP-PCR. Research on FA's impacts on DNA microarrays is still limited, necessitating further optimization in exploring the characteristics of FA in SP-PCR according to particular application needs. We immobilized DNA on a chip using a crosslinker and generated DNA microarrays through bridge amplification based on FA denaturation on our automated reaction device. We optimized the denaturation and hybridization parameters of FA, achieving a maximum cluster density of 2.83 × 104 colonies/mm2. Compared to high-temperature denaturation, FA denaturation required a lower template concentration and milder reaction conditions and produced higher cluster density, demonstrating that FA effectively improves hybridization rates on surfaces. Regarding the immobilized DNA stability, the FA group exhibited a 45% loss of DNA, resulting in a 15% higher DNA retention rate compared to the high-temperature group, indicating that FA can better maintain DNA stability. Our study suggests that using FA improves the immobilized DNA stability and amplification efficiency in SP-PCR.

Rotational Symmetry Effects on Multibody Lateral Interactions between Co-Adsorbates at Heterogeneous Interfaces.

Heterogeneous interfaces are critical in a wide range of applications, and their material properties can be tuned via changes in the coverage and configuration of chemical adsorbates. However, the tunability of such adlayers is limited by a lack of knowledge surrounding the impact of adsorbate internal structure and rotational symmetry on lateral interactions between coadsorbates. Using density functional theory (DFT) and cluster expansions, we systematically determine the impacts of rotational symmetry on lateral interactions between coadsorbates as a function of DFT functional, adsorbate type, metal type, and cluster configuration. Results indicate that the rotational symmetry effects can be nearly exclusively partitioned into the shortest 2-body clusters. By electronic analysis, the nature and strength of such effects on the lateral interactions are attributed to a balance of repulsive and attractive electrostatic interactions that are dependent on the adsorbate and metal types. Taken together, our characterization of the impacts of adsorbate internal structure and rotational symmetry on lateral interactions enables improved accuracy within multiscale modeling of multibody adsorbates at heterogeneous interfaces.

A method for chromatin domain partitioning based on hypergraph clustering

For many years, multi-scale models of chromatin domains, such as A/B compartments, sub-compartments, topologically associated domains (TADs), sub-TADs, and loops have been popular. However, existing methods can only identify structures at a single scale and cannot partition multi-scale structures. In this paper, we proposed a method (TORNADOES) for chromatin domain partitioning based on hypergraph clustering. First, we use a density clustering algorithm to identify TADs at different scales based on Hi-C data with different resolutions. Then, by combining ChIP-seq data features and TAD results at different scales, we generate a hypergraph based on these TADs. Finally, we partition the chromatin domain structure at different scales, including A/B, A1, A2, B1, B2, and B3 based on the Laplacian matrix feature of the hypergraph. Similarity comparison experiments and ChIP-seq signal enrichment analysis are performed on the A/B region and sub-TAD levels, respectively, demonstrating that our method can identify chromatin domains with distinct features and provide a deeper understanding of the organizational patterns and functional differences in TADs at the genomic hierarchical structure. Comparative analysis of multiple cell line data shows that TORNADOES can better classify different numbers and types of compartments by changing the factors ChIP-seq data and clustering number used to characterize TAD compared to other methods. Source code for the TORNADOES method can be found at https://github.com/ghaiyan/TORNADOES.

SAD-ARGRU: A Metro Passenger Flow Prediction Model for Graph Residual Gated Recurrent Networks

This paper proposes a graph residual gated recurrent network subway passenger flow prediction model considering the flat-peak characteristics, which firstly proposes the use of an adaptive density clustering method, which is capable of dynamically dividing the flat-peak time period of subway passenger flow. Secondly, this paper proposes graph residual gated recurrent network, which uses a graph convolutional network fused with a residual network and combined with a gated recurrent network, to simultaneously learn the temporal and spatial characteristics of passenger flow. Finally, this paper proposes to use the spatial attention mechanism to learn the spatial features around the subway stations, construct the spatial local feature components, and fully learn the spatial features around the stations to realize the local quantization of the spatial features around the subway stations. The experimental results show that the graph residual gated recurrent network considering the flat-peak characteristics can effectively improve the prediction performance of the model, and the method proposed in this paper has the highest prediction accuracy when compared with the traditional prediction model.

Understanding Research Dynamics in Job Stress

The objective of this analysis is to conduct a bibliometric analysis to map the research on work stress. This will be done by employing quantitative methods such as network, overlay, and density cluster models to analyze the research on work stress over a span of 11 years. Data are extracted from the scientific literature in the Science Direct database using the keyword "job stress," in which a total of 180 relevant articles were identified. These articles were analyzed using algorithms in Vosviewer to identify common topics in stress research. The analysis revealed that work-related stress studies frequently discuss 39 key topics, as determined by researchers. Through bibliometric analysis utilizing a network visualization model, it has been determined that there are six recurring clusters of keywords. The initial cluster, highlighted in red, comprises a total of 11 keywords. The terms "job stress" and "job satisfaction" have been the primary focus of research in the field of job stress analysis from 2013 to 2023. In response to the analysis of overlay visualization, the keyword "job stress" has been frequently examined from the past to the present. According to the density visualization analysis, the keywords "job stress" and "job satisfaction" exhibit the most intense and prominent colors, suggesting that numerous researchers have conducted studies on job stress.

Back-to-back high category atmospheric river landfalls occur more often on the west coast of the United States

The catastrophic December 2022-January 2023 nine atmospheric rivers in California underscore the urgent need to better understand such high-risk weather extremes. Here we applied a machine learning clustering tool to understand the activity of atmospheric river clusters. Reanalysis results show that clusters with high density, that is the time fraction under atmospheric river conditions within a cluster, exhibit more frequent high-category atmospheric rivers, alongside an increased likelihood for extreme precipitation and severe land surface responses. The key circulation patterns of atmospheric river clusters are primarily attributed to subseasonal variability. Furthermore, the occurrence and density of atmospheric river clusters are modulated by the daily variability of the geopotential height field. Climate model projections suggest that atmospheric river clusters with higher density and higher categories will be more frequent as warming level increases. Our findings emphasize the important role of atmospheric river clusters in the development of climate adaptation and resilience strategies.