Static Networks Research Articles

Curriculum negative mining for temporal networks.

Multiscale Static and Dynamic Brain Functional Network Analysis Reveals Aberrant Connectivity Patterns in Preschool Children with Autism Spectrum Disorder

ABSTRACT Stochastic violence refers to seemingly unpredictable but statistically probable acts of political violence that emerge from structural, ideological, and networked interactions. This paper presents an integrated computational framework to enhance the prediction of such violence, combining Bayesian inference, Markov chains, and network analysis. Unlike prior models that either focus on static network structures or broad probabilistic estimates, this study introduces a stepwise approach that incorporates extremist network identification, message content analysis, agent interactions, and dynamic probability updating. By examining the interplay between extremist networks and law enforcement responses, we highlight how state interventions can either mitigate or inadvertently escalate violence. A case study of the January 6th Capitol riot illustrates the practical application of this framework, demonstrating how extremist discourse, online mobilization, and enforcement responses shaped the progression of violence. Our findings suggest that integrating computational modeling with policy-oriented interventions can potentially improve early-warning systems and provide a more holistic approach to mitigating political violence in an era of increasing polarization.

Large Language Models (LLMs) have shown immense human-like capabilities for reasoning and generating digital content. However, their ability to freely sense, interact, and actuate the physical domain remains significantly limited due to three fundamental challenges: (1) physical environments require specialized sensors for different tasks, yet deploying dedicated sensors for each application is impractical; (2) events and objects of interest are often localized to small areas within large spaces, making them difficult to detect with static sensor networks; and (3) foundation models need flexible actuation capabilities to meaningfully interact with the physical world. To bridge this gap, we introduce EmbodiedFly, an embodied LLM agent combining a foundation model pipeline with a reconfigurable drone platform to observe, understand, and interact with the physical world. Our co-design approach features 1) a FM orchestration framework connecting multiple LLMs, VLMs, and an open-set object detection model; 2) a novel image segmentation technique that identifies task-relevant areas; and 3) a custom drone platform that autonomously reconfigures with appropriate sensors and actuators based on commands from the FM orchestration framework. Through real-world deployments, we demonstrate that EmbodiedFly completes diverse physical tasks with up to \(85\% \) higher success rates compared to traditional approaches leveraging static deployments.

Abstract Road networks face increasing disruptions, yet vulnerability assessment methods either oversimplify traffic dynamics or require extensive computational simulations. This research introduces a novel approach integrating traffic simulation, graph theory, and machine learning for efficient and accurate vulnerability assessment. Analysis across numerous disruption scenarios showed that static weighting is inadequate for capturing traffic redistribution effects. In contrast, dynamic weighting aligns strongly with simulation results but was computationally infeasible. To overcome this limitation, a specialized multilayer perceptron artificial neural network (ANN) model was developed with a dual‐pathway architecture and a novel automated closure propagation algorithm, separating static network attributes from spatial relationships. This surrogate model generates predictions significantly faster than traffic simulations, enabling comprehensive vulnerability analyses, previously deemed impractical. Testing across diverse disruption scales demonstrated surrogate effectiveness and limitations. This research presents a transferable and scalable methodology for constructing simulation‐informed ANN surrogate models, providing practical deployment guidance for informed resilient transportation network planning.

Community structures are fundamental in understanding the structure and functionality of complex networks. Different optimization algorithms, including both single-objective and multi-objective approaches, have been employed to address the challenge of community detection. Recently, multi-objective evolutionary algorithms (MOEAs) have attracted many researchers to identify communities in static networks. Many algorithms have been proposed to find a solution that achieves a trade-off between exploring new areas of the solution space and improving the quality of existing solutions. In this trade-off is crucial; whereas exploitation improves existing solutions, it may fail to find better solutions from insufficiently explored regions of the solution space. Therefore, mutation in evolutionary algorithms greatly impacts community detection within social networks. Conventional mutation methods usually tend to apply too much randomness, which results in convergence being less precise about finding a suitable optimum solution. This paper introduces a new mutation called community strength enhancement (CSE) to enhance the search efficiency of the Multi-Objective Evolutionary Algorithm with Decomposition (MOEA/D) and speed up the convergence of the suggested algorithm. Moreover, the proposed algorithm overcomes the limitations of traditional MOEA/D by accurately and effectively identifying communities across a wide range of social networks. The enhanced algorithm was evaluated on two groups of datasets (twenty synthetic and four real-world) using normalized mutual information (NMI) and modularity (Q) across five baseline models. Integrating the CSE mutation strategy led to significant improvements in performance, particularly under high mixing parameters and in large-scale networks, as evidenced by increased NMI and modularity scores

Covalent adaptable networks (CANs) unite recyclability and mechanical stability by leveraging externally triggered bond exchange. However, a significant challenge remains in achieving a sharp reversible transition between dynamic and static network states, which broadly limits their applicability. This work introduces a merocyanine photoacid for precise and reversible spatiotemporal control over dynamic bond exchange in thiol-ene photopolymers. When exposed to visible light, the photoacid's activated spiropyran form enables dynamic network rearrangements through acid-catalyzed transesterification. Switching off the light yields the deactivated merocyanine form and a creep-resistant polymer network. Stress relaxation experiments clearly demonstrate a remarkable difference in mechanical properties resulting from the photoacid's isomerization state. The rapid isomerization kinetics and fatigue resistance of this system are utilized to generate gradients of active photoacid at the micrometer level. Applied in a novel mold-free reshaping approach, bending radii are predictable based on the introduction of an empirical model. Finally, the successful fabrication of diverse microstructures via multiphoton laser writing highlights the future potential of these reversibly switchable CANs in light-controlled micromechanics.

Combining static and dynamic brain network analysis with machine learning for enhanced diagnosis of major depressive disorder.

Community detection is a crucial technique for uncovering latent network structures, analyzing group behaviors, and understanding information dissemination pathways. Existing methods predominantly rely on static graph structural features, while neglecting the intrinsic dynamic patterns of information diffusion and nonlinear attenuation within static networks. To address these limitations, we propose DAMA, a community detection model that integrates dynamic propagation-aware feature modeling with adaptive multi-hop structural aggregation. First, an Information Flow Matrix (IFM) is constructed to quantify the nonlinear attenuation of information propagation between nodes, thereby enriching static structural representations with nonlinear propagation dynamics. Second, we propose an Adaptive Sparse Sampling Module that adaptively retains influential neighbors by applying multi-level propagation thresholds, improving structural denoising and preserving essential diffusion pathways. Finally, we design a Hierarchical Multi-Hop Aggregation Framework, which employs a dual-gating mechanism to adaptively integrate neighborhood representations across multiple hops. This approach enables more expressive structural embeddings by progressively combining local and extended topological information. Experimental results demonstrate that DAMA achieves better performance in community detection tasks across multiple real-world networks and LFR-generated synthetic networks.

Abstract Social media plays an important role in today’s life. The dissemination of messages and the exchange of information within social networks have been extensively researched and analyzed. In social media, influence maximization (IM) is the problem of finding a small subset of the most influential nodes and maximizing influence over the entire network by their combined influence-spreading capability. We propose a non-dominated archived multi-objective harmony search (NAMHS) algorithm to identify influencers in social networks. This algorithm can generate a set of Pareto optimal harmony vectors for addressing the multi-objective IM problems. Our proposed algorithm is compared with two other state-of-the-art algorithms, and the performance of our algorithm is higher. We also propose a dynamic non-dominated archived multi-objective harmony search (DNAMHS) algorithm for dynamic networks and demonstrate that it performs better than the current algorithms. Moreover, the time complexity of the proposed NAMHS and DNAMHS algorithms is comparable with that of existing algorithms. The deterministic linear threshold model is used for influence propagation in both static and dynamic networks.

This study examines the use of Bayesian Networks (BNs) for Stealth Assessment (SA) in digital game‐based learning (DGBL) environments. By integrating in‐game behavior tracking with embedded assessment scores, we investigate how interactive gameplay fosters learning gains. Data were collected from 632 middle school students participating in Mission HydroSci (MHS), a first-person 3D narrative adventure designed to teach water science and scientific argumentation. Using Static Bayesian Networks (SBNs), we modeled probabilistic dependencies among various in‐game behaviors, including evidence‐based argumentation, tool usage, dialogue engagement, and spatial exploration, and corresponding learning outcomes measured via pre‐ and post-assessments. Our analysis reveals distinct behavioral profiles strongly linked to positive learning gains. In particular, behavior patterns, including repeated engagement in argumentation tasks, strategic tool usage, and goal-oriented spatial exploration, emerge as key predictors of enhanced performance. Insights from the BN analysis inform the design of more effective DGBL environments and highlight the potential for real-time, adaptive assessment mechanisms that maintain gameplay immersion. Overall, this research offers a data-driven framework for understanding and optimizing learning trajectories in DGBL, providing practical guidelines for educators and game designers to enhance digital learning interventions.

Disrupted static and dynamic small-world brain network topologies in patients with schizophrenia.

Reconfiguration of dynamic brain networks in heart failure with preserved ejection fraction: Linking neurovascular coupling and cardiac dysfunction.

Static and hard-coded layer-two network identifiers are well known to present security vulnerabilities and endanger user privacy. In this work, we introduce a new privacy attack against Wi-Fi access points listed on secondhand marketplaces. Specifically, we demonstrate the ability to remotely gather a large quantity of layer-two Wi-Fi identifiers by programmatically querying the eBay marketplace and applying state-of-the-art computer vision techniques to extract IEEE 802.11 BSSIDs from the seller's posted images of the hardware. By leveraging data from a global Wi-Fi Positioning System (WPS) that geolocates BSSIDs, we obtain the physical locations of these devices both pre- and post-sale. In addition to validating the degree to which a seller's location matches the location of the device, we examine cases of device movement–once the device is sold and then subsequently re-used in a new environment. Our work highlights a previously unrecognized privacy vulnerability and suggests, yet again, the strong need to protect layer-two network identifiers.

Static and dynamic functional connectivity of unruptured arteriovenous malformations-related epilepsy: A resting-state functional MRI study.

Guaranteeing reliable deep neural network (DNN) inference despite intermittent power is the cornerstone of enabling intelligent systems in energy-harvesting environments. Existing intermittent inference approaches support static neural networks with deterministic execution characteristics, accumulating progress across power cycles. However, dynamic neural networks adapt their structures at runtime. We observe that because intermittent inference approaches are unaware of this non-deterministic execution behavior, they suffer from incorrect progress recovery, degrading inference accuracy and performance. This work proposes non-deterministic inference progress accumulation to enable dynamic neural network inference on intermittent systems. Our middleware, NodPA, realizes this methodology by strategically selecting additional progress information to capture the non-determinism of the power-interrupted computation while preserving only the changed portions of the progress information to maintain low runtime overhead. Evaluations are conducted on a Texas Instruments device with both static and dynamic neural networks under time-varying power sources. Compared to intermittent inference approaches reliant on determinism, NodPA is less prone to inference non-termination and achieves an average inference speedup of 1.57 times without compromising accuracy, with greater improvements for highly dynamic networks under weaker power.

Modern society takes connectivity for granted, relying heavily on communication networks, both for interpersonal connection and to support critical infrastructure. As Internet- and data-driven technologies become increasingly pervasive, our dependence on fast, reliable communication will only deepen, necessitating advanced tools for optimizing network efficiency and resilience. Such optimization must account for the interplay between the static network infrastructure and the dynamic user preferences. The challenge is that while the infrastructure data is accessible to network operators, the user preferences, tied to personal mobility and communication habits, are protected by privacy laws and are thus heavily restricted. To address this, we introduce CLUSTER: an interpretable Bayesian nonparametric framework that leverages aggregate, low-resolution, unprotected data to identify user groups with correlated connection patterns. By uncovering these patterns, we show, CLUSTER offers actionable insights, from scheduling base-station activation to guiding deployment of new stations - all without compromising user privacy. CLUSTER thus offers a principled approach to extract meaningful insights from restricted data.

Despite the critical role of polybenzimidazole (PBI) in high-temperature membranes in fuel cells, its practical deployment remains hindered by inherent limitations including low molecular weight (<20 kDa), poor solution processability, and irreversible network formation upon conventional synthesis. To address these challenges, a dynamic covalent strategy was introduced, leveraging Diels–Alder (DA) chemistry for topological reconfiguration. By subjecting furan-functionalized PBI prepolymer (PBI-furan, Mn = 8.2 kDa) to bismaleimide chain extension, we achieve a fourfold increase in molecular weight (Mn = 32 kDa), yielding a reversibly crosslinked PBI-DA membrane. Compared to traditional PBI with the same molecular weight, this architecture synergistically integrates exceptional thermal stability (>450 °C onset decomposition), robust mechanical strength (tensile strength >80 MPa), suppressed phosphoric acid swelling (<10%) and elevated ionic conductivity. Crucially, the dynamic network enables cyclic reprocessability and autonomous self-healing, retaining >90% of the initial mechanical properties after three tensile cycles. Compared to static PBI networks, this system reduces irreversible chain entanglement, while maintaining performance parity. By deeply integrating dynamic covalent chemistry with PBI materials, this work not only advances the performance of high-temperature proton-exchange membranes but also establishes a novel framework for the sustainable design of green energy devices, presenting significant scientific merit and engineering application potential.

The rapid proliferation of the Internet of Things (IoT) has transformed digital ecosystems, enabling pervasive connectivity across industries, healthcare, smart homes, and financial infrastructures. However, the physical devices that underpin IoT ecosystems remain highly vulnerable, often constrained by limited computational capacity and weak native security mechanisms. Traditional perimeter-based defenses, designed for static enterprise networks, fail to address dynamic device-level threats, insider risks, and sophisticated adversarial tactics. This paper proposes a posture-centric adaptive cyber defense framework that leverages Zero Trust principles, Network Access Control (NAC), and Artificial Intelligence (AI)-driven anomaly detection to safeguard physical IoT devices. A mixed-methods approach, combining systematic literature review, architecture modeling, and simulated threat scenarios, was employed to evaluate the framework. Findings reveal that posture-based continuous validation enhances resilience by reducing insider risks, mitigating device-level compromise, and automating incident response. By shifting from perimeter-based to posture-centric defense, this study contributes a scalable, adaptive, and intelligence-driven security model essential for future-proofing IoT ecosystems.

Trust serves an important purpose in organizations composed of numerous, specialized, interdependent roles. Supporting confidence that individuals will dutifully fulfill the responsibilities of those roles without causing harm to the organization, trust enables coordinated task execution across multiple roles and facilitates information exchange among individuals by reducing cognitive resources spent verifying information accuracy and reliability. Interactions play an important role in shaping and updating trust, but the mechanisms underlying the relationship between communication networks and trust dynamics remain poorly understood. This paper addresses that gap by directly examining the coevolution of communication networks and trust. During a multi-day military training exercise, participants (n=83) from three distinct units formed a coalition organization largely focused on collecting, analyzing, and acting on information gleaned from the operating environment of roughly 10k units under their command. Over the course of the exercise, each participant provided eight ratings of trust in their own unit and their coalition partners’ units. Static and dynamic network models of the organization’s communication networks assessed whether trust is an antecedent or product of communication. Results consistently show that when individuals report elevated trust in a unit, they become more likely to form and sustain relationships to members of that unit during the next time period. They also increase their rates of communication to those unit members. However, this relationship does not work in reverse: Increased communication to a unit does not precede increased trust in that unit. These findings suggest temporal directionality in the coevolution of trust and communication.