Network Structure Research Articles

Optimization of Single-user Task Migration based on Improved DDPG

Aiming at the problems of slow convergence and unstable convergence of traditional reinforcement learning algorithms in minimizing computational cost on edge servers with random task arrivals and time-varying wireless channels, an improved DDPG algorithm (IDDPG) was proposed. The Critic network structure of DDPG was replaced by the Dueling structure, which converged faster by splitting the state value function into an advantage function and a value function. The update frequency of the Critic network was adjusted to be higher than that of the Actor-network to make the overall training more stable. The Ornstein- Uhlenbeck noise was added to the actions selected through the Actor-network to improve the algorithm exploration ability, and the action noise size was set in segments to ensure the stability of convergence. Experimental results show that, compared with other algorithms, the IDDPG algorithm can better minimize the computational cost and has a certain improvement in the convergence speed and convergence stability.

Network structure and taxonomic composition of tritrophic communities of Fagaceae, cynipid gallwasps and parasitoids in Sichuan, China

Abstract A key question in insect community ecology is whether parasitoid assemblages are structured by the food plants of their herbivore hosts. Tritrophic communities centred on oak‐feeding cynipid gallwasps are one of the best‐studied tritrophic insect communities. Previous work suggests that host plant identity is a much stronger predictor of oak–cynipid interactions than of cynipid–parasitoid interactions. However, these relationships have not been formally quantified. We reason that the potential for ‘bottom‐up’ effects should increase with host plant phylogenetic diversity. We, therefore, generated quantified interaction network data for previously unstudied tritrophic cynipid communities in Sichuan, China, where, in addition to Quercus, cynipid host plants include Castanea, Castanopsis and Lithocarpus. We characterise these communities taxonomically and compare the extent to which host plant taxonomy predicts plant–herbivore and plant–parasitoid associations. We sampled 42,620 cynipid galls of 176 morphotypes from 23 host plant species, yielding over 4500 specimens of 64 parasitoid morphospecies. Many parasitoids were identifiable to chalcidoid taxa present in other Holarctic oak cynipid communities, with the addition of Cynipencyrtus (Cynipencyrtidae). As elsewhere, Sichuan parasitoid assemblages were dominated by generalists. Gallwasp–plant interaction networks were significantly more modular than parasitoid–plant association networks. Gallwasps were significantly more specialised to host plants (i.e. had higher mean d' values) than parasitoids. Parasitoid assemblages nevertheless showed significant plant‐associated beta diversity, with a dominant turnover component. We summarise parallels between our study and other Fagaceae‐associated cynipid communities and discuss our findings in light of the processes thought to structure tritrophic interactions centred on endophytic insect herbivores.

Spatial network characteristics of carbon balance in urban agglomerations– a case study in Beijing-Tianjin-Hebei city agglomeration

The cities within city clusters have strong trade connections, resulting in spatial heterogeneity and correlation of carbon balances. Previous studies have not deeply explored the characteristics of the spatial association network (SAN) of carbon balance in urban agglomerations, particularly regarding network structure, node effects, and spatial-temporal inhomogeneities. This study investigated the spatial network characteristics of carbon balance in the Beijing-Tianjin-Hebei (BTH) region from 2000 to 2019, employing a modified gravity model and social network analysis (SNA). The findings revealed the following: 1) Carbon emissions increased by 106.42%, and carbon sinks increased by 31.06%, displaying spatial-temporal heterogeneity and forming a multi-level, multi-nodal SAN of carbon balance centered around Beijing. 2) The carbon balance was mainly influenced by spatial spillovers occurring at different nodes, typically moving from lower-tier to higher-tier nodes. 3) Cities assumed four roles: bidirectional spillover, net spillover, primary beneficiary, and agent, with these roles being dynamic. 4) The spatial correlation of carbon balance was primarily influenced by economic development (max 0.602), spatial distance (max 0.331), and per capita ecological land (max 0.445). This approach would refine carbon management policies and deepen the understanding of the SAN and its influencing factors, providing insights for optimizing carbon emission reduction management.

Mapping of the network connection between sleep quality symptoms, depression, generalized anxiety, and burnout in the general population of Peru and El Salvador

BackgroundA meta-analysis of randomized controlled trials has suggested a bidirectional relationship between sleep problems and mental health issues. Despite these findings, there is limited conclusive evidence on the relationship between sleep quality, depression, anxiety, and burnout.ObjectiveThe current study aimed to evaluate the relationships between sleep quality symptoms, anxiety, depression, and burnout in samples of adult individuals from two Latin American countries, Peru and El Salvador, through network analysis and to identify key symptoms that reinforce the correlation and intensify the syndromes.MethodsA total of 1012 individuals from El Salvador and Peru participated, with an average age of 26.5 years (SD = 9.1). Symptom networks were constructed for both countries based on data from the Jenkins Sleep Scale, Patient Health Questionnaire-2, General Anxiety Disorder-2, and a single burnout item.ResultsThe results indicated that Depressed Mood, Difficulty Falling Asleep, and Nervousness were the most central symptoms in a network in the participating countries. The strongest conditional associations were found between symptoms belonging to the same construct, which were similar in both countries. Thus, there is a relationship between Nervousness and Uncontrollable Worry, Anhedonia and Depressed Mood, and Nighttime Awakenings and Difficulty in Staying Asleep. It was observed that burnout is a bridge symptom between both countries and presents stronger conditional associations with Tiredness on Awakening, Depressed Mood, and Uncontrollable Worry. Other bridge symptoms include a Depressed Mood and Nervousness. The network structure did not differ between the participants from Peru and El Salvador.ConclusionThe networks formed by sleep quality, anxiety, depression, and burnout symptoms play a prominent role in the comorbidity of mental health problems among the general populations of Peru and El Salvador. The symptom-based analytical approach highlights the different diagnostic weights of these symptoms. Treatments or interventions should focus on identifying central and bridge symptoms.

Two-stage multi-objective evolutionary algorithm for overlapping community discovery

As one of the essential topological structures in complex networks, community structure has significant theoretical and application value and has attracted the attention of researchers in many fields. In a social network, individuals may belong to different communities simultaneously, such as a workgroup and a hobby group. Therefore, overlapping community discovery can help us understand and model the network structure of these multiple relationships more accurately. This article proposes a two-stage multi-objective evolutionary algorithm for overlapping community discovery problem. First, using the initialization method to divide the central node based on node degree, combined with the cross-mutation evolution strategy of the genome matrix, the first stage of non-overlapping community division is completed on the decomposition-based multi-objective optimization framework. Then, based on the result set of the first stage, appropriate nodes are selected from each individual’s community as the central node of the initial population in the second stage, and the fuzzy threshold is optimized through the fuzzy clustering method based on evolutionary calculation and the feedback model, to find reasonable overlapping nodes. Finally, tests are conducted on synthetic datasets and real datasets. The statistical results demonstrate that compared with other representative algorithms, this algorithm performs optimally on test instances and has better results.

Al-Based MOF-Derived Amorphous/Crystalline Heterophase Cobalt Sulfides as High-Performance Supercapacitor Materials.

Transition metal sulfides (TMSs) are promising electrode materials due to their high theoretical specific capacitance, but sluggish charge transfer kinetics and an insufficient number of active sites hamper their applications in supercapacitors. In this work, a self-sacrificial template strategy was employed to construct Al-based MOF-derived metal sulfides with an amorphous/crystalline (a/c) heterophase, in which aluminum, nitrogen, and carbon species were evenly coordinated in the amorphous phase. The metal sulfides a/c-Co(Al)S-1 and a/c-Co(Al)S-2, originating from the CAU-1 and CoAl-MOF on NF as self-sacrificial templates, were investigated as electrode materials, respectively, in which the a/c-Co(Al)S-1 showed a more excellent electrochemical performance. Through acid etching CAU-1 using Co(NO3)2 followed by sulfuration, the a/c-Co(Al)S-1 with a unique 3D network structure was constructed, whose unique architecture expanded the interfacial contact with the electrolyte and provided vast active sites, accelerating the charge transportation and ion diffusion. Notably, the a/c-Co(Al)S-1 displayed a high specific charge of 1791.8 C g-1 at 1 A g-1, satisfactory cycle stability, and good rate capability. The corresponding assembled a/c-Co(Al)S-1//AC device delivered a high energy density of 77.1 Wh kg-1 at 800 W kg-1 and good durability (87.4% capacitance retention over 10 000 cycles).

Air Traffic Control Speech Enhancement Method Based on Improved DNN-IRM

The quality of air traffic control speech is crucial. However, internal and external noise can impact air traffic control speech quality. Clear speech instructions and feedback help optimize flight processes and responses to emergencies. The traditional speech enhancement method based on a deep neural network and ideal ratio mask (DNN-IRM) is prone to distortion of the target speech in a strong noise environment. This paper introduces an air traffic control speech enhancement method based on an improved DNN-IRM. It employs LeakyReLU as an activation function to alleviate the gradient vanishing problem, improves the DNN network structure to enhance the IRM estimation capability, and adjusts the IRM weights to reduce noise interference in the target speech. The experimental results show that, compared with other methods, this method improves the perceptual evaluation of speech quality (PESQ), short-term objective intelligibility (STOI), scale-invariant signal-to-noise ratio (SI-SNR), and speech spectrogram clarity. In addition, we use this method to enhance real air traffic control speech, and the speech quality is also improved.

Lung X-ray image segmentation algorithm based on Multihead Self-Attention Mechanism (MSAG) optimizing Unet networks

In this paper, the Multihead Self-Attention Mechanism (MSAG) is used to optimize the Unet network for accurate segmentation of lung X-ray images. By introducing the MSAG module, the ability of the Unet network to capture global and local correlations is enhanced, which effectively improves the accuracy of the segmentation results. The introduction of the multi-head self-attention mechanism enables the network to have more powerful modelling and generalization capabilities, and can process various types of lung X-ray images stably and efficiently. The dataset is divided into training, validation and test sets according to the ratio of 4:3:3. The loss gradually converges during the training process, and the model gradually learns the data features and patterns, and the gap between them and the real labels is gradually reduced. The performance on the validation set is good and no over-fitting occurs, demonstrating the ability to generalize on unseen data. The evaluation metrics on the test set show an IoU of 0.85, a Dice of 0.92, and an Accuracy of 0.88, proving that the model can accurately extract lung features for segmentation. This study has achieved satisfactory results in the field of medical images by optimizing the network structure and introducing new techniques, which are of positive significance for improving the accuracy and efficiency of lung X-ray image segmentation.

Single Image Super-Resolution via Wide-Activation Feature Distillation Network.

Feature extraction plays a pivotal role in the context of single image super-resolution. Nonetheless, relying on a single feature extraction method often undermines the full potential of feature representation, hampering the model's overall performance. To tackle this issue, this study introduces the wide-activation feature distillation network (WFDN), which realizes single image super-resolution through dual-path learning. Initially, a dual-path parallel network structure is employed, utilizing a residual network as the backbone and incorporating global residual connections to enhance feature exploitation and expedite network convergence. Subsequently, a feature distillation block is adopted, characterized by fast training speed and a low parameter count. Simultaneously, a wide-activation mechanism is integrated to further enhance the representational capacity of high-frequency features. Lastly, a gated fusion mechanism is introduced to weight the fusion of feature information extracted from the dual branches. This mechanism enhances reconstruction performance while mitigating information redundancy. Extensive experiments demonstrate that the proposed algorithm achieves stable and superior results compared to the state-of-the-art methods, as evidenced by quantitative evaluation metrics tests conducted on four benchmark datasets. Furthermore, our WFDN excels in reconstructing images with richer detailed textures, more realistic lines, and clearer structures, affirming its exceptional superiority and robustness.

Insight of Effect of Silver Powders on Silver–Silicon Contacts of Crystalline Silicon Solar Cells through Analysis of Glass Structural Decay

Optimally designed silver–silicon (Ag–Si) contacts, based on silver (Ag) paste screen printing and the rapid sintering process, are widely utilized in mainstream crystalline silicon solar cells, enabling to reduce contact resistance and high photo‐conversion efficiency. Ag powder is the largest conductive component in Ag paste, but its impact on the formation of Ag–Si contacts lacks in‐depth research and comprehensive understanding. In this work, it is demonstrated that the oxygen adsorption capacity of silver powders affects the formation of Ag–Si contacts by enhancing the reactivity of Ag powders with glass frits, and has a significant effect on the performance of devices. The generation of Ag2O introduced by surface‐absorbed oxygen promotes the transformation of [TeO4] triangular bipyramidal into [TeO3] triangular pyramid, thus destroying the network structure of the glass frits. This structural decay accelerates the spreading of glass frits in the bulk, causing the glass frits to come into contact with SiNx layers earlier and facilitating the formation of the Ag–Si contacts. In this work, not only a complete picture of the interaction between Ag powder and glass frits but also a new direction is provided to optimize the Ag–Si contacts.

The effect of information integration on team communication in a simulated submarine control room task

Submarine control rooms are characterised by dedicated individual roles for information types (e.g. Sonar operator processes sound energy), with individuals verbally reporting the information that they receive to other team members to help resolve uncertainty in the operational environment (low information integration). We compared this work design with one that ensured critical information was more readily available to all team members (high information integration). We used the Event Analysis of Systemic Teamwork (EAST) method to analyse task, information, and social networks for novice teams operating within a simulated submarine control room under low versus high information integration. Integration impacted team member centrality (importance relative to other operators) and the nature of information shared. Team members with greater centrality reported higher workload. Higher integration across consoles altered how team members interacted and their relative status, the information shared, and how workload was distributed. However, overall network structures remained intact.

Quantifying the uncertainty of LLM hallucination spreading in complex adaptive social networks

Large language models (LLMs) are becoming a significant source of content generation in social networks, which is a typical complex adaptive system (CAS). However, due to their hallucinatory nature, LLMs produce false information that can spread through social networks, which will impact the stability of the whole society. The uncertainty of LLMs false information spread within social networks is attributable to the diversity of individual behaviors, intricate interconnectivity, and dynamic network structures. Quantifying the uncertainty of false information spread by LLMs in social networks is beneficial for preemptively devising strategies to defend against threats. To address these challenges, we propose an LLMs hallucination-aware dynamic modeling method via agent-based probability distributions, spread popularity and community affiliation, to quantify the uncertain spreading of LLMs hallucination in social networks. We set up the node attributes and behaviors in the model based on real-world data. For evaluation, we consider the spreaders, informed people, discerning and unwilling non-spreaders as indicators, and quantified the spreading under different LLMs task situations, such as QA, dialogue, and summarization, as well as LLMs versions. Furthermore, we conduct experiments using real-world LLM hallucination data combined with social network features to ensure the validity of the proposed quantifying scheme.

Full micro-nanofiber aerogel with three-dimensional fluffy network structure for rapid liquid absorption and uniform liquid retention in diapers

Diapers are essential for infants and patients with urinary incontinence. The super-absorbent polymer (SAP) in the core layer of diapers suffers from slow liquid absorption, fault formation, and reverse osmosis. Therefore, it is desirable to construct a core layer that can integrate rapid liquid absorption, uniform retention, and anti-seepage qualities. In this study, a diaper core layer and a three-dimensional fluffy network structure with a wetting gradient was constructed via one-step solution blow spinning (SBS). By regulating the content of the glutaraldehyde (GA) cross-linking agent, absorbent and insoluble fibers could be obtained. The ratio of super-absorbent fibers (SAF) to PAN-PVP micro-/nano fibers in the three-layer fiber structure and the gram weight of the core layer were adjusted. A wetting gradient was established when the weight of the adjusted gradient core layer was 2/3 that of the commercial core layer. The amount of reverse osmosis was reduced by 0.25 g, and the liquid absorption and retention were 30 times and 28.5 times, respectively, two and 3.5 times higher than those of commercial diapers. The SAF prepared by SBS could improve the core layer's liquid absorption rate, retention ratio, and comfort. The fibers are applicable in diapers, sanitary napkins, and other disposable products.

Parallel diffusion models promote high detail-fidelity photoacoustic microscopy in sparse sampling

Reconstructing sparsely sampled data is fundamental for achieving high spatiotemporal resolution photoacoustic microscopy (PAM) of microvascular morphology in vivo. Convolutional networks (CNN) and generative adversarial networks (GAN) have been introduced to high-speed PAM, but due to the use of upsampling in CNN-based networks to restore details and the instability in GAN training, they struggle to learn the entangled microvascular network structure and vascular texture features, resulting in only achieving low detail-fidelity imaging of microvascular. The diffusion models is richly sampled and can generate high-quality images, which is very helpful for the complex vascular features in PAM. Here, we propose an approach named parallel diffusion models (PDM) with parallel learning of Noise task and Image task, where the Noise task optimizes through variational lower bounds to generate microvascular structures that are visually realistic, and the Image task improves the fidelity of the generated microvascular details through image-based loss. With only 1.56% of fully sampled pixels from photoacoustic human oral data, PDM achieves an LPIPS of 0.199. Additionally, using PDM in high-speed 16x PAM prevents breathing artifacts and image distortion issues caused by low-speed sampling, reduces the standard deviation of the Row-wise Self-Correlation Coefficient, and maintains high image quality. It achieves high confidence in reconstructing detailed information from sparsely sampled data and will promote the application of reconstructed sparsely sampled data in realizing high spatiotemporal resolution PAM.

Dynamic undirected graphical models for time-varying clinical symptom and neuroimaging networks.

In this work, we propose methods to examine how the complex interrelationships between clinical symptoms and, separately, brain imaging biomarkers change over time leading up to the diagnosis of a disease in subjects with a known genetic near-certainty of disease. We propose a time-dependent undirected graphical model that ensures temporal and structural smoothness across time-specific networks to examine the trajectories of interactions between markers aligned at the time of disease onset. Specifically, we anchor subjects relative to the time of disease diagnosis (anchoring time) as in a revival process, and we estimate networks at each time point of interest relative to the anchoring time. To use all available data, we apply kernel weights to borrow information across observations that are close to the time of interest. Adaptive lasso weights are introduced to encourage temporal smoothness in edge strength, while a novel elastic fused- penalty removes spurious edges and encourages temporal smoothness in network structure. Our approach can handle practical complications such as unbalanced visit times. We conduct simulation studies to compare our approach with existing methods. We then apply our method to data from PREDICT-HD, a large prospective observational study of pre-manifest Huntington's disease (HD) patients, to identify symptom and imaging network changes that precede clinical diagnosis of HD.

Evolution and Governance of the Structure of Marine Economic Networks in China’s Coastal Areas: Based on Sea-related A-share Listed Companies

Evolution and Governance of the Structure of Marine Economic Networks in China’s Coastal Areas: Based on Sea-related A-share Listed Companies

System-level operational cyber risks identification in industrial control systems

ABSTRACT In Industrial Control Systems (ICS), where complex interdependencies abound, cyber incidents can have far-reaching consequences. Dependency modelling, a valuable technique for assessing cyber risks, aims to decipher relationships among variables. However, its effectiveness is often hampered by limited data exposure, hindering the analysis of direct and indirect impacts. We present a unique method that transforms dependency modelling data into a Bayesian Network (BN) structure and leverages causality and reasoning to extract inferences from seemingly unrelated events. Using operational ICS data, we confirm our method enables stakeholders to make better decisions about system security, stability, and reliability.

Optimization and Benefit Assessment of Production Supply Chain Networks Using Graph Neural Network Models

With the flourishing development of global economy, effective management of production supply chains is crucial for the competitiveness of enterprises. Optimizing supply chain networks can not only improve the efficiency of resource allocation but also enhance market responsiveness and systemic risk resistance. Traditional supply chain network optimization methods, focusing mostly on linear models and static analysis, fall short in addressing the growing complexity and dynamism. The emergence of Graph Neural Network (GNN) models in recent years has offered new opportunities to tackle non-linearity and structural dynamism in supply chain networks. However, existing research still faces methodological limitations in supply chain node relationship mining and benefit assessment. This study introduces an optimization and benefit assessment method for production supply chain networks based on GNNs. Firstly, by developing a node role type-aware graph neural network model, this paper achieves in-depth mining and optimization of node relationships within production supply chain networks. Secondly, a hierarchical factor analysis method is used to comprehensively assess the benefits of the production supply chain. This method can dynamically capture changes in node roles and relationships within the supply chain network, optimize the network structure, and provides a multidimensional, multilevel framework for benefit assessment. This study not only expands the application of GNN in the field of supply chain management but also provides a new analytical tool for the comprehensive assessment of supply chain benefits.

Efficient leather spreading operations by dual-arm robotic systems

To achieve precise grasping and spreading of irregular sheet-like soft objects (such as leather) by robots, this study addresses several challenges, including the irregularity of leather edges and the ambiguity of feature recognition points. To tackle these issues, this paper proposes an innovative method that involves alternately grasping the lowest point twice and using planar techniques to effectively spread the leather. We improved the YOLOV8 algorithm by incorporating the BIFPN network structure and the WIOU loss function, and trained a dedicated dataset for the lowest grasping points and planar grasping points, thereby achieving high-precision recognition. Additionally, we determined the optimal posture for grasping the lowest point and constructed an experimental platform, successfully conducting multiple rounds of leather grasping and spreading experiments with a success rate of 72%. Through an in-depth analysis of the failed experiments, this study reveals the limitations of the current methods and provides valuable guidance for future research.

Scientific Collaboration Network Structure and Connectedness Among the “Belt and Road” Countries

Scientific Collaboration Network Structure and Connectedness Among the “Belt and Road” Countries