Critical node identification in complex networks via gravity model based on steady-state restart Markov chain

The need for reliable internet connectivity in educational environments is crucial, but is often hampered by inefficient network infrastructure. This study aims to design an optimal Fiber To The Room (FTTR) network topology design based on Gigabit Passive Optical Network (GPON) at SMK NU Ma'arif Kudus, focusing on the efficiency of fiber optic cable installation routes. The research method used is engineering design with a quantitative approach, where the school architectural plan is modeled into a weighted graph. Route optimization is carried out by implementing the Dijkstra Algorithm to find the shortest path from the center node (ODC) to all node termination points (ODP). While node I (ODC) is designated as the starting node because it functions as the network distribution center. The calculation process is carried out by determining the minimum distance from the starting node to all destination nodes (ODP). The calculation results show that the shortest path is divided into two main routes, namely I to C to B to A to D to E and I to F to G to H. The selection of this node is proven to be able to produce a more efficient total distance compared to direct paths in several network segments. Based on these results, it can be concluded that Dijkstra's algorithm is effective in fiber optic network planning because it can optimally determine the path with the minimum distance. The application of this method is expected to assist in decision-making regarding fiber optic network infrastructure planning, making it more efficient and applicable for implementation in school environments.

Connectivity is a fundamental concept in network science, characterizing how interactions propagate through indirect pathways. While numerous connectivity metrics exist, such as shortest paths, effective resistance and minimum cut, each highlighting distinct structural features, their relationships remain largely fragmented. Here we show that these classical notions arise as limiting cases of a unified statistical-physics framework based on the random cluster (RC) model, which interprets connectivity as a principled synthesis of series and parallel transmission. By tuning its parameters, the RC model not only recovers classical connectivity measures but also extrapolates into unexplored regimes, leading to emergent notions of connectivity which yield practical tools for network learning tasks. In particular, RC connectivity naturally encodes the kinetics of growing paths, enhancing learning performance in dynamical settings such as epidemic spreading and neurodynamics. By linking structural, dynamical, and learning-based perspectives, RC connectivity establishes a general and interpretable foundation for the analysis of networked systems.

Cell geometry and mechanical stress coordinate stomatal division orientation.

Protein allostery plays essential roles in many biological processes through long-range communication between distant sites. Three-dimensional Shortest Path Map (3D SPM) graphs generated from molecular dynamics trajectories provide an effective, structure-embedded representation of protein conformational dynamics and allosteric communication pathways. However, comparisons among 3D SPM graphs have remained largely qualitative, limiting quantitative insight. Here, we address this gap by introducing an algorithm that uses the earth mover's distance (EMD) and the normalized graph Laplacian to quantitatively measure distances between 3D SPM graphs by evaluating their spatial distributions and network connectivity. This approach allows us to establish the sensitivity of computed allosteric communication networks to the choice of force field and other parameters in the 3D SPM calculations. Furthermore, the quantitative comparison of 3D SPM graphs enables distinguishing mutations that minimally perturb allosteric communication from those that substantially rewire the network. Importantly, decomposition of the EMD highlights the residues and edges that contribute the most to these shifts. This work has the potential to facilitate the analysis of high-throughput experiments regarding protein function and evolution, and to provide guidance for protein engineering.

ABSTRACT Motor learning is central to human adaptive behaviour. This study explored behavioural and neural changes during acquisition of a complex visuomotor skill (soccer juggling). We recruited 111 participants, assigning them to a 10-week juggling intervention group or a non-training control group. Using MRI and graph theory, we analysed white matter network changes at pre-test (T0), mid-term (T1), and post-test (T2). Results indicated that, relative to the control group, the intervention group exhibited a significant increase in global efficiency and a significant decrease in the shortest path length of the white matter network during the later learning stage (T1-T2). Conversely, during the early learning stage (T0-T1), the intervention group demonstrated no significant changes in global efficiency and shortest path length. Furthermore, the control group demonstrated a significant decrease in nodal efficiency of ORBmid.L at T0-T2 and T1-T2, whereas the intervention group showed no significant changes in this marker. Additionally, the intervention group demonstrated increased nodal degree centrality in SPG.R at T0-T1 and T0-T2, whereas controls exhibited significant declines. These findings demonstrate that extended soccer juggling training drives changes in network topology involving frontal and parietal white matter regions, particularly during late-stage skill consolidation. The study advances current knowledge by identifying temporally distinct phases of neuroplasticity, emphasising the critical role of sustained practice in optimising brain network efficiency for complex motor skill mastery.

A novel neural network model with SAGPooling graph decodes changes in cognitive trajectories.

Default mode network hub disruption links problematic use of the Internet to brain network disorganization.

To address the challenges of labor-intensive operations, high safety risks, and low operational efficiency in the installation of hanging columns within tunnels, we propose a robotics-based mechanized and automated solution. First, integrated operational equipment is designed to perform hoisting, gripping, lifting, and installation. Next, the robot linkage coordinate system is established via the modified D-H parameter method to obtain a forward kinematic model and simulate the workspace. To improve the accuracy of solving the inverse kinematic solution, a hybrid approach combining algebraic methods and genetic algorithms is proposed, and the optimal inverse solution is selected on the basis of the shortest path. Furthermore, to avoid impact and vibration while achieving optimal efficiency, a time-optimal trajectory based on seventh-degree polynomial interpolation is presented. Finally, experiments validation was conducted. The results show that the robot's workspace fully covers the required operational range. The proposed method achieves high accuracy, with orientation accuracy better than 10-6 and position accuracy better than 10 × 10-3. The motion of the joint is smooth and shock-free. The experiments confirm the equipment's capability to efficiently complete installation tasks, verifying the feasibility of the proposed solution. The research results provide a theoretical foundation for equipment design and offer important references for subsequent improvements.

Building fire key factors are the fundamental control variables that govern both the initiation of fires and dynamics of propagation. The accurate identification of key factors in building fires is crucial for enhancing the effectiveness of fire prevention strategies. To improve the accuracy of key factor identification in building fires, a novel K-shell Entropy Gravity (KEG) algorithm that integrates multiple topological metrics is proposed in this study. First, a complex network is constructed to characterize the relationships among accident factors, where nodes represent influencing factors and edges denote their co-occurrence in fire incidents. Subsequently, considering the positional importance and core connectivity of nodes, the information influence and irreplaceability of nodes, as well as the collaborative coupling and nonlinear characteristic among multiple indicators, a composite attribute integrating K-shell value, information entropy difference, and total shortest path length is developed to quantify node importance, thereby capturing both the local coreness and the global influence of nodes within the network. Then, these metrics are incorporated into an established gravity-based model to comprehensively assess the influential scope of each node, and the results are employed to identify the key factors. Finally, the proposed method is compared with baseline methods based on the Susceptible–Infected–Recovered (SIR) model and network robustness evaluation using the California Building Fire Dataset (2012–2024). In addition, a sensitivity analysis is performed to investigate how the removal of key factors affects accident propagation. To further verify the robustness of this method, fire data from Alaska are applied for comparison, and an ablation experiment is designed. The results indicate that the KEG algorithm achieves superior accuracy in identifying critical factors and offers a reliable analytical tool for developing targeted fire prevention and mitigation strategies.

Mapping relative proximity within an internalizing symptoms network.

Epitaxial LiNi1/3Mn1/3Co1/3O2 (NMC) thin films are prepared via pulsed laser deposition to model fundamental electrochemical behavior and lithium-ion transport kinetics based on different crystallographic orientations and defect types. The observed growth direction and surface termination of NMC thin films are linked to surface energy minimization, primarily via the (104) and (003) planes. The shortest diffusion path for lithium-ion transport is achieved for a film thickness of ≈15 nm via optimal (100)-oriented growth of NMC, indicating selective growth direction of NMC domains. Analysis of interfaces and local crystal structure revealed two predominant types of defects: antiphase boundaries (APBs) and twinned domains, which are strictly related to the symmetry of the layered structure and columnar epitaxial growth of NMC domains. Electrochemical testing vs Li/Li+ at charge/discharge rates from C/10 up to 6 C showed that performance is influenced by both the crystallographic orientation of lithium transport pathways and the presence of structural defects. Specifically, (104)- and (1̅08)-oriented NMC thin films with twinned microstructure exhibited stable cycling, delivering specific discharge capacities of 66.2 μA cm-2 μm-1 (141.2 mAh g-1) and 70.2 μA cm-2 μm-1 (149.4 mAh g-1) at C/10, along with apparent lithium diffusion coefficients of 7.45 × 10-15 cm2 s-1 and 7.95 × 10-15 cm2 s-1, respectively. In contrast, (003)- and (1 0 16)-oriented thin films exhibited lower apparent lithium diffusion coefficients and limited functionality due to less favorable orientations of lithium slabs, higher density of APBs, and unit cell distortion. These factors contribute to a noticeable decline in average discharge voltage at higher discharge rates across all orientations except (104). This approach reveals an intrinsic correlation between the structural properties and electrochemical response of epitaxial NMC thin films and serves as a future guideline toward high-performance NMC cathodes.

Accurate segmentation and analysis of individual trees from 3D point clouds is a crucial yet challenging task in urbanism and environmental studies. Most existing methods for tree instance segmentation suffer from either under- or over-segmentation errors, mainly due to the complex nature of the environments and the varying tree geometries. In this paper, we propose SATree, a novel structure-aware approach that directly identifies important tree structures, such as crowns and stems, from point clouds, enabling robust tree instance segmentation against tree overlaps and varying tree sizes. Our method leverages a multi-task learning framework that simultaneously performs (i) semantic segmentation to classify a point as crown , stem , or other ; (ii) heatmap prediction to assign a heat value to each point based on 2D Gaussian kernels centered at tree stem locations; (iii) offset prediction to estimate point-wise offset vectors pointing to the instance centroid. Key to our approach is the stem localization module, where we fuse the semantic and heatmap predictions to reliably localize tree stems from the network outputs. After that, we utilize a graph-based shortest path algorithm to group individual tree points by integrating the learned offset embeddings. Extensive experiments on two public forestry datasets, TreeML and ForInstance, demonstrate that SATree consistently outperforms state-of-the-art methods in terms of AP, AP 50 , and AP 25 scores, reducing significant under- or over-segmentation errors. Our research output supports downstream forestry inventory, 3D tree reconstruction, and fine-grained part segmentation of trees. We will open-source the code of SATree soon. • Structure-aware approach for 3D tree instance segmentation in large-scale forestry areas. • Fusion of crucial tree parts, including crowns and stems, to explicitly enhance segmentation robustness against crown overlaps and varying tree shapes. • Integration of heatmap prediction to learn high-response representations of major tree structures. • Precise delineation of tree boundaries through a direction-aware graph-based technique.

Shortest path problems of uncertain random networks under incomplete information environments: Models and algorithms

Optimal path for a Bi-criteria network with fractional cost functions for Automated Guided Vehicles (AGVs)

Agricultural cold chain logistics is characterized by inherent challenges-product perishability, high carbon emissions, and stringent time windows-which are further exacerbated by dynamic disruptions. Existing methods suffer from slow adaptability, unstable multi-objective convergence, and severe cold-start issues. This work falls within the broad scope of biomimetics-the science of emulating nature's time-tested strategies to solve complex engineering problems-and bio-inspired data-driven methods and their applications in engineering control, optimization, and artificial intelligence. The proposed H-MODRL framework embodies core biomimetic principles: the Genetic Algorithm (GA) mimics Darwinian natural selection and genetic inheritance, the Sparrow Search Algorithm (SSA) abstracts the cooperative foraging and anti-predation behaviors of sparrow populations in nature, and the Arrhenius-based freshness-decay model captures the biochemical kinetics governing perishable biological products. By synergistically integrating these biological evolution principles, swarm intelligence, and deep learning, the framework tackles real-world logistics complexity in a manner directly inspired by living systems. This study presents a well-organized hybrid optimization framework (H-MODRL) that couples a three-stage hybrid evolutionary mechanism, synergistically integrating heuristic warm-start, evolutionary policy guidance, and deep reinforcement learning decision-making. First, an improved genetic algorithm combined with the earliest deadline first strategy constructs a feasible initial population satisfying hard time-window constraints. Second, a large neighborhood search-enhanced chaotic sparrow search algorithm builds a high-quality elite guidance set for policy learning. Third, a physics-based multi-objective proximal policy optimization model embedded with Arrhenius equation-derived freshness-decay kinetics performs online decision-making. Experiments demonstrate that pre-computed all-pairs shortest paths and an O(1) hash-based dynamic-disruption indexing mechanism support fast online replanning. On heterogeneous simulated terrains based on real Chinese geospatial data, H-MODRL outperforms state-of-the-art algorithms across four objectives-logistics cost, carbon emissions, terminal freshness, and delivery time-while exhibiting compact, low-variance performance distributions, thereby validating its engineering robustness and practical value in complex agricultural cold chain environments.

NeuroGT: Biophysically grounded graph transformers for self-supervised representation learning of neuronal morphology.

A shortest-path and ADMM-based fluence-level optimization framework for discretized non-coplanar VMAT.

Environmentally friendly path planning has become an important topic in transportation research as concerns about carbon emissions continue to grow. This study aims to review existing research on environmentally friendly shortest path problems and to identify the current state of the art in green shortest path optimization. A Systematic Literature Review is conducted using the PRISMA guideline and supported by bibliometric analysis to examine research trends and optimization methods discussed in the literature. The review indicates that most studies focus on metaheuristic and artificial intelligence–based approaches, while deterministic methods with explicit objective prioritization receive less attention. Based on the synthesis of previous studies, this paper discusses emerging research directions and outlines a conceptual framework for priority-based multi-objective shortest path optimization. The results of this review provide a clear overview of current methods and can support future research on eco-friendly shortest path models.

This study examined the development of bilingual lexical networks in adolescence through word association task and network analysis. Participants were Chinese-English bilinguals in Grade 8 (middle school; aged 13-14years) and Grade 11 (high school; aged 16-17years). Networks were constructed based on word association responses separately for each grade and language, and structural properties of networks were computed. Results showed that from Grade 8 to Grade 11, the Chinese networks displayed increased within-group convergence while maintaining overall structural stability and small-world features. In contrast, the English networks expanded in size, with longer average shortest paths, higher local clustering, and greater modularity (Q), reflecting rapid growth and restructuring, while also exhibiting small-world features. Across grades, L1 networks remained larger and more structured than L2 networks, though the gap decreased over time, indicating increasing cross-language similarity. These findings provide new insights into bilingual lexical development during the adolescent years.