Graph anomaly detection technology has broad applications in financial fraud and risk control. However, existing graph anomaly detection methods often face significant challenges when dealing with complex and variable abnormal patterns, as anomalous nodes are often disguised and mixed with normal nodes, leading to the coexistence of homophily and heterophily in the graph domain. Recent spectral graph neural networks have made notable progress in addressing this issue; however, current techniques typically employ fixed, globally shared filters. This 'one-size-fits-all' approach can easily cause over-smoothing, erasing critical high-frequency signals needed for fraud detection, and lacks adaptive capabilities for different graph instances. To solve this problem, we propose a Multi-Head Spectral-Adaptive Graph Neural Network (MHSA-GNN). The core innovation is the design of a lightweight hypernetwork that, conditioned on a 'spectral fingerprint' containing structural statistics and Rayleigh quotient features, dynamically generates Chebyshev filter parameters tailored to each instance. This enables a customized filtering strategy for each node and its local subgraph. Additionally, to prevent mode collapse in the multi-head mechanism, we introduce a novel dual regularization strategy that combines teacher-student contrastive learning (TSC) to ensure representation accuracy and Barlow Twins diversity loss (BTD) to enforce orthogonality among heads. Extensive experiments on four real-world datasets demonstrate that our method effectively preserves high-frequency abnormal signals and significantly outperforms existing state-of-the-art methods, especially showing excellent robustness on highly heterogeneous datasets.

A novel fuzzy sliding mode control (FSMC) strategy is proposed to enhance the robustness and stability of position control for underactuated quadrotor unma

With the rapid advancement of urbanization, the frequency of urban fire incidents has been steadily increasing, posing significant challenges to public safety and sustainable urban development. Understanding the complex relationships between various urban built environments is essential for improving fire risk assessment and management. In this study, we propose an interpretable machine learning framework that incorporates Point of Interests (POIs) to enhance the accuracy and transparency of urban fire risk prediction. Using POI data from Hangzhou, a major megacity in China, we constructed and optimized three machine learning models. Among them, the Light Gradient Boosting Machine (LightGBM) demonstrated the highest prediction accuracy. To address the interpretability of the model, we employed SHapley Additive exPlanations (SHAP) to analyze the contributions of different POI types to fire incident prediction. Our findings reveal that the density of Residential areas and Life Service facilities exerts the most significant influence on fire incidence, suggesting a strong correlation between fire risk, population density, and economic activity. Moreover, interaction effects between different POI types further contribute to the complexity of urban fire risk, indicating the need for integrated and context-specific prevention strategies. This study provides a transparent, data-driven approach to assessing urban fire risk, offering valuable insights for urban planners, policymakers, and emergency service managers. By elucidating the spatial distribution patterns of fire risk in relation to POI density, our model supports more informed decision-making aimed at enhancing urban resilience and sustainability.