Network Hierarchy Research Articles

Экологический сетевой подход в приложении к биоэтическим организационным структурам

The term ‘bioethics’ (in the narrow sense) or ‘biomedical ethics’ denotes medical ethics at the modern stage of development. Bioethics is currently institutionalized and falls under the responsibility of specialized organizational structures (bioethics commissions, ‘divine committees’, etc.). The article expounds the prospects of applying network structures to institutions and organizations dealing with bioethical issues and tasks (ethical aspects of reproductive technologies, biomedical experiments, organ transplantation, and bioethical education). With the principles of decentralization (‘multiple authority’) and ecology (an integrative approach to issues under study and integrity of a bioethical expert team), network structures promote creative and effective functioning of bioethical organizations. Nonetheless, the centralized hierarchies of traditional educational and research institutions are also expected to perform essential functions. A reasonable combination of network structures and hierarchies provides the latter with a new role: the hierarchies assess the activity of emergent network structures using competent experts and provide selective support (including financing) to the most effective among them.

Controversies and progress on standardization of large-scale brain network nomenclature.

Progress in scientific disciplines is accompanied by standardization of terminology. Network neuroscience, at the level of macroscale organization of the brain, is beginning to confront the challenges associated with developing a taxonomy of its fundamental explanatory constructs. The Workgroup for HArmonized Taxonomy of NETworks (WHATNET) was formed in 2020 as an Organization for Human Brain Mapping (OHBM)-endorsed best practices committee to provide recommendations on points of consensus, identify open questions, and highlight areas of ongoing debate in the service of moving the field toward standardized reporting of network neuroscience results. The committee conducted a survey to catalog current practices in large-scale brain network nomenclature. A few well-known network names (e.g., default mode network) dominated responses to the survey, and a number of illuminating points of disagreement emerged. We summarize survey results and provide initial considerations and recommendations from the workgroup. This perspective piece includes a selective review of challenges to this enterprise, including (1) network scale, resolution, and hierarchies; (2) interindividual variability of networks; (3) dynamics and nonstationarity of networks; (4) consideration of network affiliations of subcortical structures; and (5) consideration of multimodal information. We close with minimal reporting guidelines for the cognitive and network neuroscience communities to adopt.

Analysis of the Spatial Correlation Network Structure of Agricultural Water Use Efficiency in Northwest China

There are some problems in Northwest China, such as the fragile ecological environment, poor basic conditions of agricultural production, low efficiency of agricultural water use, and difficulty in clarifying the path of agricultural water use efficiency (AWUE) improvement. Based on the superefficiency data envelopment analysis (DEA) model, this study increases the amount of ‘green water’ resources in the agricultural water consumption index of the input index and increases the wastewater pollutants (total chemical oxygen demand emissions; total ammonia nitrogen emissions) in the undesired output index to measure the AWUE in the northwest region. Based on the calculation results of AWUE, combined with the modified gravity model, the connection strength of AWUE between any two provinces in Northwest China is calculated, and the spatial structure and network characteristics of AWUE in this area are analysed via the social network analysis (SNA) method. The results show that the average AWUE in 2020 is nearly two times higher than that in 2011. From the situation of the northwest provinces, the average AWUE of the five provinces is in the order of Qinghai > Shaanxi > Gansu > Ningxia > Xinjiang. The AWUE value, total population, real GDP and per capita GDP of the capital cities of the five provinces in Northwest China are the key factors for the improvement of the connection intensity of AWUE. From 2011 to 2020, the connection intensity and closeness of AWUE in Northwest China increased, and there was a clear network hierarchy. The improvement in overall AWUE in the region is mainly due to the radiation and driving effect of the central province on other provinces. Based on this, the study proposes policy recommendations for gradually realising the AWUE improvement path of the central province (Shaanxi; Gansu)—cooperation circle (Shaanxi–Ningxia; Gansu–Qinghai)—the whole region. The results provide theoretical support and a quantitative basis for optimising the spatial pattern of agricultural water resources and improving AWUE in Northwest China.

Construction of Economic Management Performance Model of Mining Enterprises under the Background of Supply-side Reform

The proposal of supply-side structural reform measures has ushered in new development opportunities for mining enterprises in the bottleneck period. As a result, a model for evaluating the economic management performance of mining firms at the new performance level must be established. This paper suggests an assessment model for the economic management of mining firms against the backdrop of supply-side reform in light of this. This study reasonably incorporated financial and non-financial performance indicators, constructed the Economic Management Performance (EMP) evaluation index system of mining enterprises, and created an economic management performance evaluation model of mining enterprises based on the BP neural network and analytic hierarchy process. The study selected the relevant data of five mining companies A, B, C, D, and E from 2017 to 2022 as the research object, verified the effectiveness of the model, and analyzed the performance evaluation results of the companies. The research results show that the model constructed in this study can evaluate the economic management performance level of enterprises within a reasonable range (the mean relative error is 1.98%). Since 2017, the comprehensive performance level of these five mining companies has gradually declined. But thanks to the supply-side reform, the comprehensive performance has gradually recovered after 2022 and among the five mining companies, company A has always been at the performance level way ahead. Overall, the model developed in this research has strong operability and practicability and can be utilized more effectively to forecast the mining industry’s potential for future growth.

Exploring the role of texture features in deep convolutional neural networks: Insights from Portilla-Simoncelli statistics

It is well-understood that the performance of Deep Convolutional Neural Networks (DCNNs) in image recognition tasks is influenced not only by shape but also by texture information. Despite this, understanding the internal representations of DCNNs remains a challenging task. This study employs a simplified version of the Portilla-Simoncelli Statistics, termed “minPS,” to explore how texture information is represented in a pre-trained VGG network. Using minPS features extracted from texture images, we perform a sparse regression on the activations across various channels in VGG layers. Our findings reveal that channels in the early to middle layers of the VGG network can be effectively described by minPS features. Additionally, we observe that the explanatory power of minPS sub-groups evolves as one ascends the network hierarchy. Specifically, sub-groups termed Linear Cross Scale (LCS) and Energy Cross Scale (ECS) exhibit weak explanatory power for VGG channels. To investigate the relationship further, we compare the original texture images with their synthesized counterparts, generated using VGG, in terms of minPS features. Our results indicate that the absence of certain minPS features suggests their non-utilization in VGG’s internal representations.

Exploring cross-video matching for few-shot video classification via dual-hierarchy graph neural network learning

Few-shot video classification methods aim to recognize a new class with only a few training examples. Distinct from previous few-shot methods, we explicitly consider the relations in cross-video domains and take full advantage of the cross-video frame matching in a hierarchy learning fashion. In this paper, we propose a Dual-Hierarchy Graph Neural Network to realize comprehensive cross-video frame matching and video relation modeling. In the first hierarchy of the graph neural network, we build a Cross-video Frame Matching Graph to extract robust frame-level features via accumulating information across frames sampled from both query and support videos. Then, frame representations are accumulated to obtain the video-level features. In the second hierarchy of the graph neural network, we construct a Video Relation Graph by taking the video-level features as nodes, which can adaptively learn positive relations between query and support videos. We get the predicted label of the query video through the matching learning of edges connecting video nodes. We evaluate the model on three benchmarks: HMDB51, Kinetics, and UCF101. Extensive experiments on benchmark datasets demonstrate that our model significantly improves few-shot video classification across a wide range of competitive baselines and showcases the strong generalization of our framework. The source code and models will be publicly available at https://github.com/JiaMingZhong2621/DHGNN.

Spatial network structure and driving factors of human settlements in three Northeastern provinces of China

IntroductionUrban human settlements' spatial network structures have emerged as crucial determinants impacting their health and sustainability. Understanding the influencing factors is pivotal for enhancing these settlements. This study focuses on 34 prefecture-level cities in Northeastern China from 2005 to 2020. It employs a modified gravitational model to establish spatial relationships among urban human settlements. Social network analysis techniques, including modularity and the quadratic assignment procedure (QAP) regression model, are introduced to analyze the network's characteristics and driving factors.MethodsA modified gravitational model is applied to create the spatial association network of urban human settlements. Social network analysis tools, along with modularity and the QAP regression model, are utilized to investigate the network's attributes and influencing elements. The study evaluates the evolution of spatial correlation, network cohesion, hierarchy, and efficiency.ResultsThroughout the study period, spatial correlation among urban human settlements in Northeastern China progressively intensified. However, the network exhibited relatively low density (0.217675), implying limited interconnectivity among cities. The average network hierarchy was 0.178225, indicating the need for optimization, while the average network efficiency was 0.714025, reflecting fewer redundant relationships. The analysis reveals the emergence of a polycentric network pattern with core and sub-core cities like Shenyang, Dalian, Changchun, Daqing, and Harbin. The urban network configuration has largely stabilized. The spatial association network showcases the intertwining of "small groups" and community organizations. Geographic proximity and merit-based linkages govern feature flow. Measures such as breaking administrative barriers, reducing flow time and distance, boosting resident income, and increasing government investment are identified to foster balanced network development and structural optimization.DiscussionThe research underscores the increasing spatial correlation and evolving network pattern among urban human settlements in Northeastern China. Despite the observed strengthening correlation, challenges related to network cohesion and hierarchy persist. The formation of a polycentric network signifies positive progress in urban development. The study highlights the importance of proximity and merit-based connections for feature flow. The proposed measures offer pathways to enhance network development and optimize structure, promoting holistic urban settlement growth and sustainability.

Classification of broadband network devices using text mining technique

The Broadband Internet industry is highly competitive, with service providers investing heavily in network development to meet customer demands and competing on pricing. Effective cost management is crucial for profitability in this market. This work proposes a model for classifying broadband network devices based on text mining techniques applied to a device list from a leading broadband network company in Thailand. The device descriptions are used to generate a feature vector, which is then employed by a classification algorithm to categorize devices into core, access, and last mile hierarchies. Various algorithms including decision tree, naïve Bayes, Bayesian network, k-nearest neighbor, support vector machine, and deep neural network are compared, with support vector machine achieving the highest accuracy of 90.35%. The results are visualized to provide insights into network hierarchy, device replacement dates, and budget requirements, enabling support for cost management, budget planning, maintenance, and investment decision-making. The methodology outline includes,•Obtaining a device list from a major broadband network company and extracting device descriptions through text mining and generating a feature vector.•Using a support vector machine for classification and comparing algorithm performances.•Visualizing the results for actionable insights in cost management, budget planning, and investment decisions.

Assessing the Stability of the Core/Periphery Structure and Mobility in the Post-2008 Global Crisis Era

How did the hierarchy of the world-system adapt to the impact of the 2008–09 global economic crisis? How did a country's position in the world-system influence their upward mobility during the crisis? This paper investigates the core/periphery hierarchy of the global trade network before and after the 2008–09 crisis. The central argument posits that the global trade network follows a core/periphery hierarchy in relation to the new international division of labor (NIDL) in the twenty-first century, and a country's placement within that hierarchy had a varying effect on their upward mobility following the 2008–09 crisis. Utilizing social network analysis of 191 countries engaged in global trade, I discover that the core/periphery structure remained unchanged after the 2008–09 global financial crisis, although many countries in intermediate positions experienced upward shifts. However, not all countries were able to achieve upward mobility, indicating that only a few semi-peripheral and peripheral countries were better positioned to improve their status compared to most non-core countries.

Research On Bidding for The Olympic Games Based on Meta-Network and ISM Model

In view of the dwindling interest of countries to bid for hosting the Olympic Games, this paper proposes two models to assess the impact and feasibility of hosting the Olympic Games. The first model is an Olympic Impact D&A System Meta-Network model based on 20 indicators in four categories including economy, society, politics and environment, and the correlations between the indicators are classified and determined by Principal Component Analysis (PCA). The second model is based on Interpretive Structural Modeling (ISM), which proposes four new hosting methods and analyzes their centrality, causality and hierarchy in the new network. The results show that fixed location in developed countries and annual smaller game in developing countries are the two most advantageous methods.

An Approach for Quick Method of Estimating Passenger And Aircraft Demand for Feeder Air Services in India

The recent and the proposed entry of new airlines both in the ‘no-frills’ and in the "full service" categories has completely changed the aviation scenario in India. Obviously this explosive increase in capacity will have a profound effect on the economy of the airlines both in the short term as well as in the long-term period. A growth of over 25% per annum in passenger traffic may not be sustainable for long; thus galloping increase in capacity may result in mergers and strategic alliances of airlines on one side and closing down of some of the airlines on the other side. A point, which is to be noted is that, most of the new entrants also operate in the major trunk routes and serve the metros and the major cities. What appears to have not been tapped is the equally large potential in the regional and short haul sectors, which serve major, and small city pairs and also act as feeders to the major hubs.Realizing the importance of this, the government in its new civil aviation policy has given thrust to regional/ short-haul air travel. In this paper, an attempt is made to review the hierarchy of airline networks in India with respect to trunk, major and feeder routes. The air passenger traffic carried in terms of revenue passenger kilometers (RPK) is compared with upper class rail traffic. The present status of feeder airline services in India is outlined. It is pertinent to study the passenger traffic demand in these feeder routes as compared to upper class rail traffic as some percentage of upper class rail passengers are willing to shift to air travel due to marginal difference between airfare and upper class rail fare and time factor. The future RPKs are forecasted based on the actual RPKs data and the estimated RPKs of earlier market survey. A suitable aircraft size for airline operations in these feeder routes is arrived at based on existing air traffic data by a simplified approach from the basic principles of airline business. A macroscopic measure is used to determine the aircraft demand by assuming average block speed and utilization for each aircraft type. A sensitivity analysis was carried out for various levels of traffic growth keeping in view the sharp increase of air travel in recent years and further growth in air traffic fuelled by feeder sector as the hinterland opens up industry, services and tourism.

Implementation of a Novel Audio Network Protocol

Recently, the rapid advancement of the IT industry has resulted in significant changes in audio-system configurations; particularly, the audio over internet protocol (AoIP) network-based audio-transmission technology has received favourable evaluations in this field. Applying the AoIP in a certain section of the multiple-cable zone is advantageous because the installation cost is lower than that for the existing systems, and the original sound is transmitted without any distortion. The existing AoIP-based technology, however, cannot control the audio-signal characteristics of every device and can only transmit multiple audio signals through a network. In this paper, the proposed Audio Network & Control Hierarchy Over peer-to-peer (Anchor) system enables all audio equipment to send and receive signals via a data network, and the receiving device can mix the signals of different IPs. Accordingly, it was possible to improve the system-application flexibility by simplifying the audio-system configuration. The research results confirmed that the received audio signals from different IPs were received, mixed, and output without errors. It is expected that Anchor will become a standard for audio-network protocols.

Uncovering the relational upgrading of the iron product network in China

Iron is a crucial resource for societal and economic progress. How can the China's iron industry, having grown in size and gained an export advantage, seek partners in the industry chain to develop better growth potential? We propose "relational upgrading" across multi-hierarchy. Using a Three-Stage Least Squares approach, we analyze the impact of three key network centralities of iron products on industrial upgrading, based on 1995–2020 HS-4 coded product export data. Results show: China's iron product network is multi-hierarchy. Relational upgrading transcends export advantage enhancement and reflects jumps to higher at multi-hierarchy. Policymakers can evaluate industry competitiveness by monitoring the iron product network's hierarchy of upgrades. Degree and betweenness centrality have a significant "inverted-U" effect on the iron industry's upgrading, while closeness has a hindering effect. Upgrading relationships across multi-hierarchy consolidates the network centrality status of iron products. Some policies for sustainable development as indicated by the results are given.

HKGCL: Hierarchical graph contrastive learning for multi-domain recommendation over knowledge graph

Multi-domain recommendation (MDR) aims to improve the recommendation performance in all target domains simultaneously by leveraging rich data from relevant domains. However, conventional approaches either only rely on sparse interactions in each separate scenario, or simply combine multiple shared recommendation solutions, resulting in MDR systems still suffering from severely low recommendation accuracy. To tackle the above issues, we propose a novel recommendation model called Hierarchical Graph Contrastive Learning for Multi-Domain Recommendation over Knowledge Graph (HKGCL). Different from previous MDR models, our HKGCL can treat each relevant domain as a hierarchy in the interaction network. Based on the LightGCN aggregation strategy, a hierarchical message passing mechanism over the knowledge graph is designed to aggregate hierarchical knowledge representations for users and items. Then, a hierarchical node dropping scheme is proposed as our data augmentation technique on different hierarchical views to obtain more self-supervised semantic signals. Subsequently, three graph contrastive learning tasks in multiple and separate domains are proposed to explore domain-shared and domain-specific preference features for target users, respectively. Additionally, a novel Multi-domain Bayesian Personalized Ranking (MBPR) approach is proposed to assist in training our multi-task learning framework. Extensive experiments conducted on two real-world datasets demonstrate the consistent superiority of our proposed HKGCL over various state-of-the-art baselines, and also verify that the HKGCL can achieve strong performance on sparse interaction scenarios.

Atypical functional connectivity hierarchy in Rolandic epilepsy

Functional connectivity hierarchy is an important principle in the process of brain functional organization and an important feature reflecting brain development. However, atypical brain network hierarchy organization in Rolandic epilepsy have not been systematically investigated. We examined connectivity alteration with age and its relation to epileptic incidence, cognition, or underlying genetic factors in 162 cases of Rolandic epilepsy and 117 typically developing children, by measuring fMRI multi-axis functional connectivity gradients. Rolandic epilepsy is characterized by contracting and slowing expansion of the functional connectivity gradients, highlighting the atypical age-related change of the connectivity hierarchy in segregation properties. The gradient alterations are relevant to seizure incidence, cognition, and connectivity deficit, and development-associated genetic basis. Collectively, our approach provides converging evidence for atypical connectivity hierarchy as a system-level substrate of Rolandic epilepsy, suggesting this is a disorder of information processing across multiple functional domains, and established a framework for large-scale brain hierarchical research.

Savitar: an intelligent sign language translation approach for deafness and dysphonia in the COVID-19 era

PurposeIn the COVID-19 era, sign language (SL) translation has gained attention in online learning, which evaluates the physical gestures of each student and bridges the communication gap between dysphonia and hearing people. The purpose of this paper is to devote the alignment between SL sequence and nature language sequence with high translation performance.Design/methodology/approachSL can be characterized as joint/bone location information in two-dimensional space over time, forming skeleton sequences. To encode joint, bone and their motion information, we propose a multistream hierarchy network (MHN) along with a vocab prediction network (VPN) and a joint network (JN) with the recurrent neural network transducer. The JN is used to concatenate the sequences encoded by the MHN and VPN and learn their sequence alignments.FindingsWe verify the effectiveness of the proposed approach and provide experimental results on three large-scale datasets, which show that translation accuracy is 94.96, 54.52, and 92.88 per cent, and the inference time is 18 and 1.7 times faster than listen-attend-spell network (LAS) and visual hierarchy to lexical sequence network (H2SNet) , respectively.Originality/valueIn this paper, we propose a novel framework that can fuse multimodal input (i.e. joint, bone and their motion stream) and align input streams with nature language. Moreover, the provided framework is improved by the different properties of MHN, VPN and JN. Experimental results on the three datasets demonstrate that our approaches outperform the state-of-the-art methods in terms of translation accuracy and speed.

A network approach to atomic spectra

Network science provides a universal framework for modeling complex systems, contrasting the reductionist approach generally adopted in physics. In a prototypical study, we utilize network models created from spectroscopic data of atoms to predict microscopic properties of the underlying physical system. For simple atoms such as helium, an a posteriori inspection of spectroscopic network communities reveals the emergence of quantum numbers and symmetries. For more complex atoms such as thorium, finer network hierarchies suggest additional microscopic symmetries or configurations. Furthermore, link prediction in spectroscopic networks yields a quantitative ranking of yet unknown atomic transitions, offering opportunities to discover new spectral lines in a well-controlled manner. Our work promotes a genuine bi-directional exchange of methodology between network science and physics, and presents new perspectives for the study of atomic spectra.

Convexity Hierarchies in Grid Networks

Several algorithms for path planning in grid networks rely on graph decomposition to reduce the search space size; either by constructing a search data structure on the components, or by using component information for A* guidance. The focus is usually on obtaining components of roughly equal size with few boundary nodes each. In this paper, we consider the problem of splitting a graph into convex components. A convex component is characterized by the property that for all pairs of its members, the shortest path between them is also contained in it. Thus, given a source node, a target node, and a (small) convex component that contains both of them, path planning can be restricted to this component without compromising optimality. We prove that it is NP-hard to find a balanced node separator that splits a given graph into convex components. However, we also present and evaluate heuristics for (hierarchical) convex decomposition of grid networks that perform well across various benchmarks. Moreover, we describe how existing path planning methods can benefit from the computation of convex components. As one main outcome, we show that contraction hierarchies become up to an order of magnitude faster on large grids when the contraction order is derived from a convex graph decomposition.

Meta Learning in Decentralized Neural Networks: Towards More General AI

Meta-learning usually refers to a learning algorithm that learns from other learning algorithms. The problem of uncertainty in the predictions of neural networks shows that the world is only partially predictable and a learned neural network cannot generalize to its ever-changing surrounding environments. Therefore, the question is how a predictive model can represent multiple predictions simultaneously. We aim to provide a fundamental understanding of learning to learn in the contents of Decentralized Neural Networks (Decentralized NNs) and we believe this is one of the most important questions and prerequisites to building an autonomous intelligence machine. To this end, we shall demonstrate several pieces of evidence for tackling the problems above with Meta Learning in Decentralized NNs. In particular, we will present three different approaches to building such a decentralized learning system: (1) learning from many replica neural networks, (2) building the hierarchy of neural networks for different functions, and (3) leveraging different modality experts to learn cross-modal representations.

The disruption of functional connectome gradient revealing networks imbalance in pediatric bipolar disorder

ObjectivePediatric bipolar disorder (PBD) is a psychiatric disorder marked by alteration of brain networks. However, the understanding of these alterations in topological organization still unclear. This study aims to leverage the functional connectome gradient to examine changes in functional network hierarchy in PBD. MethodConnectome gradients were used to scrutinize the differences between functional gradient map in PBD patients (n = 68, aged 11 to 18) and healthy controls (HC, n = 37, aged 11 to 18). The association between regional altered gradient scores and clinical factors was examined. We further used Neurosynth to determine the correlation of the cognitive terms with the PBD principal gradient changes. ResultsGlobal topographic alterations were exhibited in the connectome gradient in PBD patients, involving gradient variance, explanation ratio, gradient range, and gradient dispersion in the principal gradient. Regionally, PBD patients revealed that the default mode network (DMN) held the most majority of the brain areas with higher gradient scores, whereas a higher proportion of brain regions with lower gradient scores in the sensorimotor network (SMN). These regional gradient differences exhibited significant correlation with clinical features and meta-analysis terms including cognitive behavior and sensory processing. ConclusionFunctional connectome gradient presents a thorough investigation of large-scale networks hierarchy in PBD patients. This exhibited excessive segregation between DMN and SMN supports the theory of imbalance in top-down control and bottom-up in PBD and provides a possible biomarker for diagnostic assessment.