Network Size Research Articles

Source inference for misinformation spreading on hypergraphs

Source inference aims at revealing the seed of the misinformation spreading on social networks, and attracted great attention in the field of network science and cybersecurity. Extensive real-world data analyses have certificated that individual interactions exist pairwise and higher-order interactions, and thus should be described using the hypergraph. Previous studies about the source inference algorithms are mainly focused on simple graphs (i.e., a graph only has pairwise interactions) while neglecting the higher-order interactions. In this article, we propose a dynamical message-passing (DMP) algorithm to infer the misinformation spreading on hypergraphs. As a comparison, we also extend jordan centrality (JC), betweenness centrality (BC), closeness centrality (CC), and eigenvector centrality (EIG) algorithm to hypergraphs. The results show that our proposed DMP algorithm can accurately and effectively infer the propagation source on both artificial and real-world hypergraphs compared to the centrality algorithms. Even if the increase in propagation scale renders the other centrality algorithms completely ineffective, the DMP algorithm can still distinguish the real propagation sources in the form of a rank smaller than 10% of the network size. In general, our DMP algorithm provides an effective solution for inferring the misinformation source on high-order networks.

GCN-MHSA: A novel malicious traffic detection method based on graph convolutional neural network and multi-head self-attention mechanism

With the increasing size and complexity of network, network traffic becomes more and more correlated with each other, and the traditional manner of presenting network traffic in a Euclidean structure is difficult to effectively capture the correlation information of network traffic. In contrast, graph structured data has gained much attention in recent years due to its ability to represent the correlation between different traffic flows; In addition, models and algorithms related to Graph Convolution Neural network (GCN) have been used for malicious traffic detection. However, existing GCN-based malicious traffic detection methods still suffer from incomplete description of the flow-level features of network traffic, imperfect traffic correlation establishment mechanism and failure to distinguish the importance of features during model training. Based on this, this study proposes a malicious traffic detection method called GCN-MHSA based on Graph Convolutional Neural network and Multi-Head Self-Attention mechanism. Firstly, the flow-level features of network traffic are populated and more information close to the features are selected to describe the network traffic; And then, the link homogeneity is used to establish the correlations between network traffic; Moreover, multi-head self-attention mechanism is introduced in the GCN model to provide larger weight to important features; Finally, an improved GCN is used as a deep learning model to detect malicious traffic. Extensive experimental results on three publicly available network traffic datasets and a real network traffic dataset show that the proposed GCN-MHSA method performs better than five baselines in terms of detection effect and stability, with an improvement of about 2.4% in accuracy, recall and F1-measure as well as an improvement of about 2.1% in precision.

Pipe merging for transient gas network optimization problems

In practice, transient gas transport problems frequently have to be solved for large-scale gas networks. Gas network optimization problems typically belong to the class of Mixed-Integer Nonlinear Programming Problems (MINLP). However current state-of-the-art MINLP solvers are not yet mature enough to solve large-scale real-world instances. Therefore, an established approach in practice is to solve the problems with respect to a coarser representation of the network and thereby reducing the size of the underlying model. Two well-known aggregation methods that effectively reduce the network size are parallel and serial pipe merges. However, these methods have only been studied in stationary gas transport problems so far. This paper closes this gap and presents parallel and serial pipe merging methods for transient gas transport problems. An empirical evaluation indicates that the developed methods perform very accurately on a huge set of fine-grained real-world data taken from one of the largest transmission system operators in Europe.

Robustness of higher‐order network pertinent to first‐ and higher‐order dependency

AbstractRobustness of higher‐order network (HON) with dependency is different from first‐order network (FON), as the attack orders and the number of iterations are totally different. To measure the robustness of HON with dependency, this study proposed new attack strategies on node and edge failure through measuring the size of the giant component and global efficiency. And two real‐world systems of hazardous materials transportation accident causation system and Chinese high‐speed railway system with different network size were empirically studied. Results show that HON is more fragile to robustness attacks than FON. The number of attack iterations is uncertain in HON, which depends on the attack strategies and the network structure. Meanwhile, nodes and edges with high dependency have greater impacts on network robustness, providing valuable insights into capturing critical information. This study may be helpful to design the robust system, improve the system's reliability and serve as a valuable reference for future exploration of HONs robustness.

Optimizing Windowing Techniques to Improve the Accuracy of Artificial Neural Networks in Predicting Outpatient Visits

Hospitals are healthcare institutions that play an important role in providing health services to the community. To optimize the service, hospitals need to predict the number of outpatient visits. The objectives of this research are (1) determine the effect of window size on the accuracy of predicting the number of outpatient visits, and (2) identify the best window size and accuracy of neural networks in predicting the daily number of outpatient visits. To achieve the research objectives, the following steps were undertaken: data collection of outpatient visits at RSUD dr. Soedirman Kebumen from 2018 to 2023, preprocessing, applying different window sizes, modeling neural networks, and testing by calculating the RMSE value for each window size. The test results show that the lowest RMSE for 2018 was 1.267 with a window size of 34, for 2019 was 1.262 with a window size of 34, for 2020 was 1.515 with a window size of 17, for 2021 was 1.81 with a window size of 18, for 2022 was 1.282 with a window size of 20, and for 2023 was 1.263 with a window size of 29. These window sizes indicate the cycle of outpatient visits each year. By understanding these visit cycles, the number of outpatient visits can be predicted at any time.

Sports following and social capital in the United States: Social networks and trust

Scholarship on sports following represents a burgeoning field of inquiry in the sociology of sport, yet sport following's connections to, or ability to facilitate, social capital remains underexplored. Drawing on a contemporary nationwide survey of the United States and employing the foundational theories of social capital by which we operationalize statistical analyses using regression modeling, this study investigates the nuanced contributions of sports following to various dimensions of social capital. Our analysis reveals complex relationships of sports following to social network size, community engagement, neighborhood trust, and social network variety, highlighting sports following as a significant predictor for all our operationalized social capital measures, although negatively so for neighborhood trust. Exploring the interplay between sports following, individual characteristics, and socio-political affiliations, results underscore the powerful predictive capacity of sports following even controlling for other characteristics, highlighting strong potentiality for building social capital, although along some dimensions more than others. These findings spotlight the domain of sports following as fertile ground for further investigation into its capacity to serve as a unique facilitator of social capital on the increasingly divided American socio-cultural landscape.

A deep-level decomposed model to accelerate hydraulic simulations in large water distribution networks

As the size of water distribution network (WDN) models continues to grow, developing and applying real-time models or digital twins to simulate hydraulic behaviors in large-scale WDNs is becoming increasingly challenging. The long response time incurred when performing multiple hydraulic simulations in large-scale WDNs can no longer meet the current requirements for the efficient and real-time application of WDN models. To address this issue, there is a rising interest in accelerating hydraulic calculations in WDN models by integrating new model structures with abundant computational resources and mature parallel computing frameworks. This paper presents a novel and efficient framework for steady-state hydraulic calculations, comprising a joint topology-calculation decomposition method that decomposes the hydraulic calculation process and a high-performance decomposed gradient algorithm that integrates with parallel computation. Tests in four WDNs of different sizes with 8 to 85,118 nodes demonstrate that the framework maintains high calculation accuracy consistent with EPANET and can reduce calculation time by up to 51.93 % compared to EPANET in the largest WDN model. Further investigation found that factors affecting the acceleration include the decomposition level, consistency of sub-model sizes and sub-model structures. The framework aims to help develop rapid-responding models for large-scale WDNs and improve their efficiency in integrating multiple application algorithms, thereby supporting the water supply industry in achieving more adaptive and intelligent management of large-scale WDNs.

Multi-generational dispersal and dynamic patch occupancy reveals spatial and temporal stability of seascapes

The success of non-native species (NNS) invasions depends on patterns of dispersal and connectivity, which underpin genetic diversity, population establishment and growth. In the marine environment, both global environmental change and increasing anthropogenic activity can alter hydrodynamic patterns, leading to significant inter-annual variability in dispersal pathways. Despite this, multi-generational dispersal is rarely explicitly considered in attempts to understand NNS spread or in the design of management interventions. Here, we present a novel approach to quantifying species spread that considers range expansion and network formation across time using the non-native Pacific oyster, Magallana gigas (Thunberg 1793), as a model. We combined biophysical modelling, dynamic patch occupancy models, consideration of environmental factors, and graph network theory to model multi-generational dispersal in northwest Europe over 13 generations. Results revealed that M. gigas has a capacity for rapid range expansion through the creation of an ecological network of dispersal pathways that remains stable through time. Maximum network size was achieved in four generations, after which connectivity patterns remained temporally stable. Multi-generational connectivity could therefore be divided into two periods: network growth (2000−2003) and network stability (2004–2012). Our study is the first to examine how dispersal trajectories affect the temporal stability of ecological networks across biogeographic scales, and provides an approach for the assignment of site-based prioritisation of non-native species management at different stages of the invasion timeline. More broadly, the framework we present can be applied to other fields (e.g. Marine Protected Area design, management of threatened species and species range expansion due to climate change) as a means of characterising and defining ecological network structure, functioning and stability.

Network analyses show horizontal and vertical distribution of vascular epiphytes on their hosts in a fragment of cloud forest in Central Mexico.

Cloud forests figure as one of the most diverse ecosystems, accounting not only for a high number of plant species but also with a great variety of interactions among them. A common interaction in these forests is the one between vascular epiphytes and their hosts. However, few studies have used the network approach to analyze them. Here, we analyze the horizontal and vertical structure of the vascular epiphyte - host network in a cloud forest in central Mexico. We quantified the number of epiphyte stands on each host both total and per-stratum. Complete network, group, and species metrics were estimated at both levels of analysis. The host - epiphyte networks had relatively low network size but were highly connected; moderately nested, with low specialization, and modularity; but higher vulnerability than generality, and high niche overlap. The community was composed by a high number of generalist species. To our knowledge this is the first study in which network analyses are conducted with standardized data and including all host and epiphyte species in the community. The analyses suggest that the networks are robust, and that functional redundancy might be probable, two advantageous characteristics in a very fragmented and threatened cloud forest.

Multiple anthropogenic environmental stressors structure soil bacterial diversity and community network

Microbial communities in many ecosystems are suffering a wide range of environmental stressors induced by anthropogenic perturbations. While the impacts of a single stressor have been extensively estimated in numerous studies, the responses of microbial communities to multiple environmental stressors simultaneously are still poorly understood. In the current study, we investigated the taxonomic diversity, community resistance, and co-occurrence interaction of soil bacterial communities treated with different numbers of environmental stressors by conducting 136 microcosms. We found that the richness and Shannon diversity of the soil community decreased significantly from 1430 to 6.54 in the mono-factor treatments to 920 and 5.77 in the hepta-factor treatments. The counts of nodes and edges of the soil microbial networks decreased with the increasing stressor number, potentially indicating that multiple stressors can reduce the network size. Multiple stressors increased the community resistance potential to environmental disturbance. Additionally, the network cohesion suggested that the cooperative and competitive behaviors between microorganisms were induced by multiple stressors. The observation could be potentially due to the enrichment of the generalists by multiple environmental stressors. Although only a handful of stressors were included, our study still indicated that multiple environmental stressors would lead to diversity loss via deterministic processes.

The Application of Carbon-Based Fertilizer Changed the Microbial Composition and Co-Occurrence Network Topological Properties of Vineyard Soil

Charcoal-based fertilizer could be used extensively and is environmentally friendly. An experiment was designed to investigate the effects of different charcoal-based fertilizer application methods on soil microbiology and grape quality in a vineyard to guide the cultivation of ‘Shine-Muscat’. A control treatment without fertilization and six other treatments were set up. Four treatments applied carbon-based fertilizer as a base fertilizer with or without potassium fulvic acid, a complex microbial agent, or Bacillus subtilis, and two treatments were only applied with two applications of carbon-based fertilizer or compound fertilizer during the expansion period. The results showed that the bacterial phyla were mainly Proteobacteria and Bacteroidetes. Ascomycota, Basidiomycota, and Mortierellomycota dominated the fungal community. At the genus level, the composition of fungi, compared to bacteria, varied significantly, while the dominant flora differed among fertilization practices. Application of charcoal-based fertilizer enriched beneficial microorganisms, while chemical fertilizers enriched pathogenic microorganisms. The addition of microbial fungicides and biostimulants for a period reduced the size of the microbial network, lowered positive correlations, and enhanced resistance to adverse conditions and diseases and there was no significant correlation between agronomic traits and microbial network topology. A combination of soil microbial and grape agronomic traits suggests that a charcoal-based fertilizer base, with microbial fungicides applied, is the optimal fertilization regimen for grape.

General and distinct correspondence between personality disorder and interpersonal functioning

The alternative model of personality disorders (AMPD) in DSM-5 includes interpersonal dysfunction as a core construct as a global severity dimension. However, it is less known how various interpersonal characteristics contribute to both general and distinct dimensions of personality dysfunction. In participants from community sources, we obtained responses to Levels of Personality Functioning Scale-Self Report (LPFS-SR), maladaptive traits (PID-5-BF), and social relationship patterns, including those related to close relationships and quantitative measures of network size. Canonical correlation analysis mapped conjoint associations between two sets of variables (personality scales and social relationship) and identified three distinct modes of correlation as significant. The first canonical pattern represented global dysfunction and was associated with utilitarianism, short-termed, weaker strengths, and smaller network sizes. The second canonical correlation represented externalizing traits and was associated with a larger number of relationships, higher utilitarianism, and short-term relationships with a close significant other. The third canonical correlation represented a detached, unemotional, and callous personality which corresponded with weaker relationship strength with both the mother and a close significant other. Our findings suggest that interpersonal functioning corresponding to personality dysfunction can be distinguished into both common and specific characteristics and further highlight the importance of characterizing distinct patterns within close relationships.

Trends in Post-Acute Care use in Medicare Advantage Versus Traditional Medicare: A Retrospective Cohort Analysis

ObjectivesWe sought to describe national trends in hospitalization and post-acute care utilization rates in skilled nursing facilities (SNFs) and home health (HH) for both Medicare Advantage (MA) and Traditional Medicare (TM) beneficiaries, reaching up to the COVID-19 pandemic (2015–2019). DesignRetrospective, observational using 100% sample of Medicare Provider Analysis and Review file (MedPAR), the Medicare Beneficiary Summary File, the Minimum Data Set (MDS), and the Outcome and Assessment Information Set (OASIS). Setting and ParticipantsMedicare beneficiaries aged 66 and older enrolled in MA or TM who were hospitalized and discharged alive. MethodsWe first calculated the proportions of MA and TM beneficiaries who were hospitalized and who used any post-acute care, as well as the total number of days of post-acute care used. We also calculated the size of the post-acute care network used by TM and MA beneficiaries within each hospital in our sample and the measured quality (star ratings) of the post-acute care providers used. ResultsWe found hospitalizations, SNF stays, and HH stays were all decreasing over time in both populations. Although similar proportions of MA and TM beneficiaries received SNF or HH care, MA beneficiaries received fewer days. The largest difference we found was in the number of post-acute care providers used in TM and MA, with MA using far fewer; however, quality ratings were similar among post-acute care providers used in each program. Conclusions and ImplicationsTogether, these results suggest MA beneficiaries have fewer days in post-acute care, receive care from fewer providers of similar measured quality to TM, but have a similar number of days outside the hospital or SNF in the first 100 days after hospital discharge.

Emotional support across adulthood: A 60-year study of men's social networks.

Past research shows that social networks get smaller with age. But not all types of relationships may shrink at the same rate or for similar reasons. In the present study, we used a unique data set from a sample of 235 men who were followed longitudinally for 71 years to examine how the general pattern of network shrinkage documented in previous research generalizes to the number of emotional support providers in people's networks. We additionally examined early-life predictors of the size of later-life support network. Growth curve analyses revealed that, mirroring the more general pattern of network shrinkage, emotional support networks shrink by as much as 50% between the ages of 30 and 90, reflecting an average reduction from two to one support providers. Examining the associations between prospectively collected measures of childhood family environment and later-life emotional support, we also found that men who grew up in warmer family environments had larger support networks in adulthood. In contrast, childhood family socioeconomic status was not connected to the size of emotional support networks later in life. The generalizability of this work is limited by the use of an archival all-male sample from the United States. Despite this limitation, these findings make important contributions to our understanding of adult socioemotional development and underscore the importance of prospectively collected longitudinal data in developmental research. Additional research is needed to examine the consequences of changing emotional support across the lifespan for health and well-being. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Generalized [formula omitted]-sparse algorithms for constructing fault tolerant RBF networks

In the construction process of radial basis function (RBF) networks, two common crucial issues arise: the selection of RBF centers and the effective utilization of the given source without encountering the overfitting problem. Another important issue is the fault tolerant capability. That is, when noise or faults exist in a trained network, it is crucial that the network’s performance does not undergo significant deterioration or decrease. However, without employing a fault tolerant procedure, a trained RBF network may exhibit significantly poor performance. Unfortunately, most existing algorithms are unable to simultaneously address all of the aforementioned issues. This paper proposes fault tolerant training algorithms that can simultaneously select RBF nodes and train RBF output weights. Additionally, our algorithms can directly control the number of RBF nodes in an explicit manner, eliminating the need for a time-consuming procedure to tune the regularization parameter and achieve the target RBF network size. Based on simulation results, our algorithms demonstrate improved test set performance when more RBF nodes are used, effectively utilizing the given source without encountering the overfitting problem. This paper first defines a fault tolerant objective function, which includes a term to suppress the effects of weight faults and weight noise. This term also prevents the issue of overfitting, resulting in better test set performance when more RBF nodes are utilized. With the defined objective function, the training process is designed to solve a generalized M-sparse problem by incorporating an ℓ0-norm constraint. The ℓ0-norm constraint allows us to directly and explicitly control the number of RBF nodes. To address the generalized M-sparse problem, we introduce the noise-resistant iterative hard thresholding (NR-IHT) algorithm. The convergence properties of the NR-IHT algorithm are subsequently discussed theoretically. To further enhance performance, we incorporate the momentum concept into the NR-IHT algorithm, referring to the modified version as “NR-IHT-Mom”. Simulation results show that both the NR-IHT algorithm and the NR-IHT-Mom algorithm outperform several state-of-the-art comparison algorithms.

Predictors of dementia among the oldest old: longitudinal findings from the representative “survey on quality of life and subjective well-being of the very old in North Rhine-Westphalia (NRW80+)”

Background/AimsOur current study aimed to investigate the determinants of dementia among the oldest old using longitudinal data from a representative sample covering both community-dwelling and institutionalized individuals.Methods/DesignLongitudinal representative data were taken from the “Survey on quality of life and subjective well-being of the very old in North Rhine-Westphalia (NRW80+)” that surveyed community-dwelling and institutionalized individuals aged 80 years and above (n = 1,296 observations in the analytic sample), living in North Rhine-Westphalia (most populous state of Germany). The established DemTect was used to measure cognitive impairment (i.e., probable dementia). A logistic random effects model was used to examine the determinants of probable dementia.ResultsThe mean age was 86.3 years (SD: 4.2 years). Multiple logistic regressions revealed that a higher likelihood of probable dementia was positively associated with lower education (e.g., low education compared to medium education: OR: 3.31 [95% CI: 1.10–9.98]), a smaller network size (OR: 0.87 [95% CI: 0.79–0.96]), lower health literacy (OR: 0.29 [95% CI: 0.14–0.60]), and higher functional impairment (OR: 13.45 [3.86–46.92]), whereas it was not significantly associated with sex, age, marital status, loneliness, and depressive symptoms in the total sample. Regressions stratified by sex were also reported.DiscussionOur study identified factors associated with dementia among the oldest old. This study extends current knowledge by using data from the oldest old; and by presenting findings based on longitudinal, representative data (also including individuals residing in institutionalized settings).ConclusionsEfforts to increase, among other things, formal education, network size, and health literacy may be fruitful in postponing dementia, particularly among older women. Developing health literacy programs, for example, may be beneficial to reduce the burden associated with dementia.

Intentional travel group and social network: Identification and dynamics during a pandemic

Using multiday continuous smartcard data in 2020, we investigate group-based travel in Hong Kong metro system by identifying metro riders intentionally traveling in groups (ITGs). ITGs serve as our proxies for citywide physical social interactions. Considering ITG members are interrelated through group-based trips, we construct a social network (an ITG network) formed by ITGs to explore the network properties and structures of ITG activities. Examining ITGs both before and during the COVID-19 pandemic, we measure the spatial patterns of ITGs and their dynamics across locales and over time. We find that the degree of the ITG network follows a heavy-tailed distribution. The network size and interconnections vary across time. Some ITG members are more influential vertices than others in maintaining the networks’ topological properties. We illustrate how new data and methods can be used to explore in-person interactions and social activity patterns in transit-reliant cities.

Racial and Ethnic Variations in Resilience Factors Among Sexual and Gender Minority Midlife and Older Adults.

This study explores resilience factors among sexual and gender minority (SGM) midlife and older adults, identifying historical/environmental, psychological, social, and behavioral predictors, and examining variations across racial/ethnic subgroups. By adopting a resilience-focused perspective, this research contributes to understanding strengths in the SGM community in the context of aging. Using weighted survey data from Aging with Pride: National Health, Aging, and Sexuality/Gender Study, this research investigates distinct risk and protective factors associated with resilience among SGM midlife and older adults. Subgroup variations were assessed using regression models. Factors moderated by race/ethnicity were integrated into final interaction models. Hispanic and Black individuals showed significantly higher resilience than non-Hispanic Whites. Lifetime victimization and day-to-day discrimination were negatively associated with resilience. Positive associations were found for higher income, identity affirmation, larger network size, greater social participation, increased physical activity, and sufficient food intake. Interaction models revealed nuanced patterns by subgroups; education negatively impacted resilience among Black individuals, while income positively influenced Hispanic individuals' resilience. The Other racial/ethnic group demonstrated unique associations between optimal sleep and resilience. This study shifts from a deficit-based to a resilience-focused approach among SGM midlife and older adults, revealing key strengths within diverse subgroups. Results underscore the significance of recognizing racial/ethnic differences in factors promoting resilience and posing risks for SGM midlife and older adults. Tailoring interventions to address the intersectional needs of SGM aging populations is essential for enhancing their abilities to bounce back from adverse events.

Which Social Variables Predict Diabetes Onset? Robust Findings in Two National Surveys.

The role of social factors in diabetes onset has been obscured by wide variation in their conceptualization and operationalization. We apply 3 theoretical frameworks to categorize social relationship variables along several dimensions and identify which dimension(s) are robustly associated with incident diabetes in the older adult population. The National Social Life, Health, and Aging Project (n = 2,365) and the Health and Retirement Study (n =11,824) provided longitudinal data from 57 to 90-year-old respondents over a 4- to 5-year period. Logistic regression models were used to test associations of 15 social variables measured identically in both data sets with diabetes onset measured as respondents' first report of a physician's diagnosis. In both studies, not being married, experiencing strain in a spousal relationship, and feeling lonely were associated with increased risk for diabetes onset at follow-up. Inconsistent or null findings were observed for social support, social activity, network size, number of friends and relatives, living alone, and closeness to network members. Robust findings in 2 large-scale surveys support the importance of the valence dimension (i.e., positive and negative); specifically, alleviating negative aspects of social life might more effectively reduce risk for diabetes than augmenting positive ones. Findings were not aligned with social variables differing on the subjectivity dimension (i.e., structural, functional, and qualitative aspects of social connections). Future work needs consistent conceptualization and measurement of social factors to correctly identify and categorize risk factors for diabetes onset and other health conditions in older adults.

Graph neural networks as strategic transport modelling alternative ‐ A proof of concept for a surrogate

AbstractPractical applications of graph neural networks (GNNs) in transportation are still a niche field. There exists a significant overlap between the potential of GNNs and the issues in strategic transport modelling. However, it is not clear whether GNN surrogates can overcome (some of) the prevalent issues. Investigation of such a surrogate will show their advantages and the disadvantages, especially throwing light on their potential to replace complex transport modelling approaches in the future, such as the agent‐based models. In this direction, as a pioneer work, this paper studies the plausibility of developing a GNN surrogate for the classical four‐step approach, one of the established strategic transport modelling approaches. A formal definition of the surrogate is presented, and an augmented data generation procedure is introduced. The network of the Greater Munich metropolitan region is used for the necessary data generation. The experimental results show that GNNs have the potential to act as transport planning surrogates and the deeper GNNs perform better than their shallow counterparts. Nevertheless, as expected, they suffer performance degradation with an increase in network size. Future research should dive deeper into formulating new GNN approaches, which are able to generalize to arbitrary large networks.