Search Results (33)

Processing-in-memory and in-memory computing (PIM/IMC) are increasingly explored to mitigate the von Neumann data-movement bottleneck that limits deep neural network (DNN) performance and energy efficiency. Progress, however, remains fragmented across device substrates, architectural prototypes, mapping and scheduling methods, compiler toolchains, and benchmarking practices, making results hard to compare and slowing deployment. This survey synthesizes developments from 2019–2025 along four coupled axes: (i) memory substrates and architectural design, (ii) mapping, partitioning, and scheduling, including learning- and graph-based strategies, (iii) compilers and end-to-end deployment flows, and (iv) benchmarking datasets, metrics, and reporting norms. Drawing on over twenty representative platforms spanning static random-access memory (SRAM) and dynamic random-access memory (DRAM), emerging non-volatile, capacitive, and photonic substrates, we clarify the trade-offs separating analog/charge-domain IMC from digital SRAM/DRAM-centric PIM, including reported peaks up to 600 TOPS/W and 1.5 TOPS/mm². We organize mapping frameworks into a unified reference taxonomy, identify recurrent evaluation pitfalls that undermine reproducibility, and highlight persistent gaps in training support, robustness under non-idealities, and coverage of large-scale GNN workloads. Finally, we outline a five-phase roadmap from benchmark standardization to industrial validation toward compiler-integrated, GNN-informed PIM/IMC systems validated on production-scale workloads. Full article

Movie genre classification is a significant challenge in narrative analysis, as traditional methods often fail to capture complex structural relationships within movie stories. This study introduces the Intra-Cluster Weighted Movie Network (ICWMN), a novel framework designed to improve classification by using intra-movie relationships through Graph Neural Networks (GNNs). We constructed a large-scale dataset of 1631 movie character networks using an automated pipeline comprising web scraping, regular expressions, and fine-tuned BERT models for entity recognition. To address the computational limitations of fully connected models, we partition ICWMN into clusters and establish edges only between the k-most similar nodes using the K-Nearest Neighbor algorithm and various distance measures, such as the Laplacian and NetLSD. XGBoost is applied to optimize high-dimensional node feature vectors. Experimental results demonstrate outstanding performance, with the Graph Attention Network (GAT) emerging as the top-performing architecture, resulting in classification accuracies that peak at $95.00\%$ on our 1631-movie dataset and an exceptional $97.30\%$ on the 773-movie Moviegalaxies dataset. These findings confirm that prioritizing spectral properties and cluster-based network topologies significantly improve the precision and stability of genre classification compared to state-of-the-art methods. Full article

Protein–ligand binding affinity prediction underpins structure-based drug discovery, yet random partitions of public benchmarks overestimate generalisation due to protein-family and ligand leakage, and the marginal value of explicit pocket-geometry descriptors over atom-level graph neural networks remains unclear. We computed a 59-dimensional discrete differential geometry descriptor on the ligand-aware solvent-excluded surface of 3285 PDBBind v2020 complexes, combining curvature distributions, the leading sixteen Laplace–Beltrami eigenvalues and a ten-point heat-kernel signature, and evaluated it in gradient-boosted tree pipelines across progressively stricter split regimes and two leak-proof external benchmarks, together with four mechanistically distinct injection strategies in a SchNet-style graph neural network. The descriptor lifted Pearson correlations by 0.111 on cluster-disjoint testing, 0.258 on LP-PDBBind DataSAIL S2 and 0.365 on CASF-2016, while in isolation reaching 0.456 to 0.594 on external benchmarks, on a par with X-Score and AutoDock Vina (version 1.2). TreeSHAP attribution localised the dominant signal to the heat-kernel signature. The four graph neural network injection strategies produced no statistically significant lift, indicating that distance-based message passing on atomic coordinates already captures much of the geometric content. Pocket-surface discrete differential geometry, therefore, offers an interpretable, leakage-robust and lightweight feature class for early-stage virtual screening, and motivates hybrid mesh-to-atom architectures. Full article

Weak-bus identification is a key task for online security assessment, preventive control, maintenance verification, and resilience-oriented dispatch of interconnected power grids. In large-scale grids, conventional full-graph graph neural networks preserve the complete network topology but may become inefficient when many operating scenarios must be screened repeatedly. Direct graph partitioning improves computational tractability, but it may cut tie-line channels and weaken the boundary evidence that determines cross-area risk propagation. To address this trade-off, this paper proposes a boundary-compensated partition-based parallel graph neural network for weak-bus identification. The method first constructs a scenario-aware weighted power-grid graph and divides it into electrically coherent subgraphs under coupling-strength and partition-size constraints. Local graph encoders are then executed in parallel to learn intra-partition vulnerability representations. A boundary compensation module further restores cross-partition information by weighting tie-line neighbors according to electrical coupling, branch loading, and cross-area association. Standardized partition scores are finally fused into a whole-grid weak-bus ranking, and a composite learning objective jointly considers node-score regression, boundary consistency, and pairwise ranking stability. The method is evaluated on the IEEE 57-bus benchmark with mechanism-based node and branch vulnerability labels. Compared with the original full-graph GNN, the proposed method reduces the mean square error from 0.0359 to 0.0147, improves the Spearman rank coefficient from 0.248 to 0.446, and increases Hit@10 from 30% to 70%. Topological interpretation further shows that the identified weak buses are concentrated around high-risk branches such as 8-12, 12-14, 0-14, and 7-8, indicating that the proposed framework captures local aggregation, boundary transmission, and corridor-driven vulnerability propagation. The IEEE 57-bus benchmark is used as a focused validation case because it provides aligned node- and branch-level vulnerability evidence for evaluating weak-bus ranking behavior. Because the available aligned vulnerability evidence is concentrated in this medium-scale benchmark, the results should be interpreted as a focused validation of the proposed ranking mechanism rather than as a complete large-system scalability study. Full article

The rapid growth of maritime communication networks introduces significant challenges to routing optimization, arising from large-scale network topologies, highly dynamic node mobility, and stringent real-time communication requirements. Conventional routing algorithms often exhibit limited scalability and poor adaptability when facing frequent topology variations. The routing problem is modeled as a multi-agent distributed decision-making process, where each node acts as an autonomous agent. In this paper, we propose a graph-partitioned hierarchical graph representation learning framework (GP-HGRL) for scalable and continual routing optimization in dynamic maritime networks. By explicitly modeling the network as a time-evolving graph, GP-HGRL first partitions the global topology into topology-aware subgraphs, enabling distributed learning and inference with reduced computational complexity. A hierarchical graph neural network architecture is then developed to jointly capture intra-subgraph local structures and inter-subgraph global dependencies, producing topology-aware embeddings for routing decision-making. Based on the learned representations, a deep reinforcement learning policy is employed to perform distributed next-hop routing decisions. To effectively handle topology dynamics induced by node mobility and link variations, we further introduce a continual graph learning mechanism that selectively updates representations and routing policies only within affected subgraphs, thereby avoiding costly global retraining and preserving routing stability. Extensive simulations demonstrate that GP-HGRL consistently outperforms shortest-path routing and existing reinforcement learning-based approaches in terms of packet delivery ratio, retransmission rate, packet loss, and training efficiency under various network loads and dynamic conditions. Full article

In this study, a special version of the Flexible Job Shop Scheduling Problem with equally and consistently batching constraints (hereafter called ECBFJSP) is considered, which involves multiple aspects of coordination, such as machine selection, process sorting, and batch splitting, which is highly complex and places strict demands on the optimization strategy. To effectively meet this challenge, this study constructs a dual-action deep reinforcement learning algorithm framework based on the Enhanced Heterogeneous Graph Neural Network (EHGNN). First, an enhanced heterogeneous graph and EHGNN model for the ECBFJSP is innovatively proposed. By integrating multi-dimensional node features such as work order priority, machine tool processing capability, and process constraints, dynamic feature aggregation of various types of information is achieved with the help of GATs and GRUs. The model can output context-aware representations containing global resource constraints, greatly improving the joint optimization efficiency of job scheduling and batch partitioning and significantly enhancing the adaptability of the dual-action decision framework to the complexity of the ECBFJSP. At the decision-making mechanism level, this study designed a dual-action decision space of process sequencing–machine selection action and batch partitioning action and used the DAPPO algorithm to collaboratively optimize the dual-action strategy to ensure the stability and efficiency of the decision-making process. The experimental data results show that compared with traditional algorithms, the proposed intelligent decision framework performs better in scheduling quality when solving the ECBFJSP, which fully verifies the significant effectiveness and practicality of the framework in solving the ECBFJSP. Full article

Major depressive disorder (MDD) is a prevalent and severe mental disorder, and EEG-based automated detection has become a promising approach for auxiliary screening diagnosis. In this work, we propose a novel multiscale spatiotemporal graph neural network for MDD detection from multichannel EEG signals. Specifically, a left–right hemispheric partitioning prior is used to encode brain functional organization. Based on this partitioning, adaptive graphs are then constructed and graph message passing is performed to model intra-hemispheric interactions. The approach not only incorporates brain functional organization into the learning process but also enhances the extraction of discriminative features related to depressive brain dynamics. The proposed method was validated in a cross-subject scenario on a private resting-state EEG dataset including 54 adult participants (27 MDD patients and 27 healthy controls; age range: 27–48 years). Experimental results on the dataset achieve an accuracy of 92.21%, surpassing the baseline models. Meanwhile, ablation experiments demonstrate the effectiveness of our proposed method. Full article

The increasing complexity of modern gate-level circuits significantly degrades the efficiency of existing Hardware Trojan detection methods. Community partitioning is an efficient structural decomposition technique to address efficiency and scalability issues, yet current community-based detection schemes rely primarily on undirected graph modeling. To address these issues, we propose an improved structure-aware community detection method for gate-level netlists, aiming to enhance the detection and localization capabilities of small-scale Hardware Trojans. First, an expanded dataset with structural diversity of clean and Trojan-inserted circuits is constructed by extending Trust-Hub benchmark circuits. Then, a directed and resolution-adaptive Louvain community detection algorithm is proposed—by introducing directed modularity, resolution parameters, and logic-gate semantic weighting, fine-grained community partitioning is achieved. On this basis, topological, functional, and anomaly features are extracted from community subgraphs, and a detection framework is built by combining graph neural networks and traditional detection models. All experiments are conducted on a unified platform equipped with an Intel (R) Core (TM) i7-10750H processor and an NVIDIA GeForce RTX 2060 GPU. Experimental results show that compared with configurations using the original Louvain partitioning and traditional features, the proposed method achieves significant improvements in both detection accuracy and localization capability. After introducing the improved community partitioning and feature design, the optimal model (CommunityGAT) yields a 3.3% increase in TPR and a 10.8% increase in ALC, verifying the method’s effectiveness in detecting small-scale concealed Trojans. Full article

Robotic polishing in CAD-free industrial settings relies on point-cloud data, yet noise and non-uniform sampling often compromise kinematic feasibility and finishing quality. This paper proposes an adaptive motion planning approach with explicit kinematic constraints. A downsampling–clustering–mapping-back strategy is first employed for rapid workpiece extraction. Subsequently, an improved supervoxel representation and attributed adjacency graph (AAG) are developed, utilizing a multi-objective energy formulation to partition sub-regions that satisfy geometric consistency and kinematic reachability. To handle point-cloud noise, a lightweight neural network predicts scanning directions and step-distance coefficients, followed by thick-slice serpentine path generation. Finally, closed-loop verification ensures safety through inverse-kinematics and safety-margin checks. Experimental results demonstrate consistent sub-micron finishing quality, with Ra ≈ 0.6 μm on complex mold surfaces. Moreover, the proposed pipeline achieves a 7.5× preprocessing speedup, completing workpiece extraction in 1.14 s for a 237,640-point scan, and improves kinematic feasibility to 100% IK success while reducing the mean TCP normal deviation by ~76% compared with a PCA-based baseline. Full article

Community detection in graphs can be viewed as the estimation of a partition map that remains stable under admissible perturbations of graph topology and node attributes. While modern graph neural networks (GNNs) achieve strong empirical accuracy, they often exhibit severe assignment drift under minor perturbations, leading to illusory community structures. In this work, we propose DISPEL-GNN, a stability-aware graph learning framework that integrates spectral operator regularization, Bayesian uncertainty modeling, and risk-aware dynamic attention for perturbation-bounded community detection. The model explicitly constrains graph operators through uniform spectral norm bounds, high-frequency energy suppression, and commutator alignment while dynamically modulating message passing based on node-level spectral risk and epistemic uncertainty. We further formalize instability via assignment of drift functional and establish perturbation bounds linking drift to operator norms and spectral gaps, complemented by a PAC-Bayesian generalization guarantee. Extensive experiments on real-world benchmarks including Cora, Citeseer, Pubmed, Cora-Full, and DBLP demonstrate that DISPEL-GNN consistently reduces assignment drift by 18–35% under feature noise and edge perturbations while improving clustering quality with up to +3.0 NMI and +0.04 ARI compared to strong baselines such as GAT and Bayesian GNNs. The normalized mutual information (NMI), adjusted Rand index (ARI), and PAC-Bayesian (PAC) constraints serve as evaluative and theoretical instruments in this study. Additional studies on synthetic graphs with controlled spectral gaps confirm that the proposed method maintains stable community assignments in low-gap regimes where classical spectral and GNN-based methods degrade sharply. These results establish DISPEL-GNN as a mathematically grounded and practically effective framework for robust and interpretable community detection. A metric-wise dominance analysis shows that DISPEL-GNN achieves metric-wise dominance across most accuracy and robustness criteria, with minor tradeoffs in modularity on selected datasets. These results indicate that explicitly modeling stability and uncertainty provides a principled pathway toward reliable and interpretable community detection in noisy graph environments. Full article

This study aims to investigate the computation of the partition dimension of various anti-myocardial infarction drugs, a graph-theoretical invariant of molecular graphs representing these drugs, for understanding and computationally characterizing structural properties of molecular networks. To improve the computational modeling of this topological invariant, advanced neural network techniques, specifically graph neural networks (GNNs) and deep neural networks (DNNs), are adopted. The GNN captures topological and molecular connection features from the molecular graph structures, which are then input into the DNN model. The DNN further processes these features to estimate the partition dimension, evaluating training performance, performing regression analysis, and producing error histograms. The model’s predictions are validated against reference values. Moreover, by analyzing the role that symmetry plays in determining the calculation of partition dimension, studying how the GNN takes advantage of permutation invariance concept related to symmetry principles to provide the DNN with symmetry-invariant features, and relating the degree of molecular symmetry to the predictive model’s accuracy and performance, its structural interpretation rather than direct chemical behavior. This dual-model approach permits a comprehensive evaluation of the model’s effectiveness in apprehending the structural characteristics of molecular graphs derived from drug molecules. The results are explicated in detail, focused on prediction accuracy, error distributions, and regression results. Moreover, this graph-theoretical metric analysis of partition dimension supports structure-based drug analysis and computational modeling, rather than direct prediction of pharmacokinetic properties, by integrating artificial neural network applications into pharmaceutical research. Full article

Aiming at the challenges of complex spatial-temporal correlation and strong nonlinearity in the power prediction of large-scale wind farm clusters, this study proposes a short-term wind power prediction method that combines a dynamic graph structure and a Kolmogorov–Arnold Network (KAN) enhanced neural network. Firstly, a spectral embedding fuzzy C-means (FCM) cluster partition method combining geographic location and numerical weather prediction (NWP) is proposed to solve the problem of insufficient spatio-temporal representation ability of traditional methods. Secondly, a dynamic directed graph construction mechanism based on a stacked wind direction matrix and wind speed mutual information is designed to describe the directional correlation between stations with the evolution of meteorological conditions. Finally, a prediction model of dynamic graph convolution and Transformer based on KAN enhancement (DGTK-Net) is constructed to improve the fitting ability of complex nonlinear relationships. Based on the cluster data of 31 wind farms in Gansu Province of China and the cluster data of 70 wind farms in Inner Mongolia, a case study is carried out. The results show that the proposed model is significantly better than the comparison methods in terms of key evaluation indicators, and the root mean square error is reduced by about 1.16% on average. This method provides a prediction tool that can adapt to time and space changes for engineering practice, which is helpful to improve the wind power consumption capacity and operation economy of the power grid. Full article

The Border Gateway Protocol (BGP) is the core protocol for inter-domain routing on the Internet. However, due to its lack of built-in security authentication mechanisms, BGP is highly vulnerable to misconfigurations or malicious route announcements, which can lead to severe incidents such as route hijacking and information leakage. Existing detection methods face two major bottlenecks: First, as the scale of Autonomous System (AS)-level topology continues to grow, conventional graph neural networks struggle to meet the demands of computational resources and latency. Second, the observational data provided by current monitoring systems are inherently localized. To address these challenges, this paper proposes a Graph Learning-driven framework for BGP Anomaly Detection, named GLBAD. The core design of GLBAD comprises three components: First, to handle BGP’s large-scale network topology, we propose a graph partition method to perform a dedicated topological partitioning on the BGP network. Second, to overcome the limitation of localized observational data, we design a graph autoencoder-based approach for adaptive graph learning, enabling topology inference. Finally, integrating the above components, we develop a comprehensive BGP anomaly detection system to achieve real-time and accurate anomaly detection. We evaluate our approach on 20 real-world BGP anomaly events. Experimental results demonstrate that the proposed GLBAD effectively detects anomalies with less time consumption while achieving a lower false positive rate. Full article

Real-world systems frequently exhibit hierarchical multipartite graph structures, yet existing graph neural network (GNN) approaches lack systematic frameworks for hyperparameter optimization in heterogeneous multi-level architectures, limiting their practical applicability. This study proposes a Bayesian optimization framework specifically designed for heterogeneous GNNs operating on three-level graph structures, addressing the computational challenges of configuring partition-aware architecture. Four GNN architectures—Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), Graph Isomorphism Networks (GINs), and GraphSAGE—were systematically evaluated using Gaussian Process-based Bayesian hyperparameter optimization with inter-partition message-passing mechanisms. The framework was validated on the TIMSS 2023 dataset (10,000 students, 789 schools, 25 countries), demonstrating that Bayesian-optimized GraphSAGE achieved the highest explained variance (R² = 0.6187, RMSE = 71.73, MAE = 64.32) compared to seven baseline methods. Bayesian optimization substantially improved model performance, revealing that two-layer architectures optimally capture cross-partition dependencies in three-level structures. GNNExplainer was used to identify the most influential student-level features learned by the model, providing explanatory insight into how the model represents individual characteristics. The optimization framework is broadly applicable to other heterogeneous and multi-level graph settings; however, the empirical findings, such as the optimal architecture depth, are specific to hierarchical graphs with structural properties like the TIMSS topology. Full article

The power system, as a critical national infrastructure, faces stealthy and persistent intrusions from Advanced Persistent Threat (APT) attacks. These attack chains span multiple stages and components, while heterogeneous data sources lack unified semantics, limiting the interpretability of current detection methods. To address this, we combine the W3C PROV-DM standard with power-specific semantics to map generic provenance data into standardized provenance graphs. On this basis, we propose a graph neural network framework that jointly models temporal dependencies and structural features. The framework constructs unified provenance graphs with snapshot partitioning, applies Functional Time Encoding (FTE) for temporal modeling, and employs a graph attention autoencoder with node masking and edge reconstruction to enhance feature representations. Through pooling, graph-level embeddings are obtained for downstream detection. Experiments on two public datasets show that our method outperforms baselines across multiple metrics and exhibits clear inter-class separability. In the context of scarce power-domain APT data, this study improves model applicability and interpretability, and it provides a practical path for provenance graph-based intelligent detection in critical infrastructure protection. Full article