Multi-resolution Score-Based Variational Graphical Diffusion for Causal Disaster System Modeling and Inference

2025-04-05

AI Summary

Modeling multi-source, heterogeneous, and multi-resolution observational data in disaster systems remains challenging due to the need to jointly capture static causal chains and dynamic evolutionary dependencies. To address this, we propose a causal-driven multi-resolution fractional diffusion framework. Our method introduces a novel parent-guided child-evolution causal scoring mechanism—the first of its kind—and integrates stochastic differential equations, variational graph neural networks, score matching, and sequence prediction to jointly model physical consistency and temporal dynamics. The framework supports irregular spatial-temporal sampling and inference under weak prior knowledge. Evaluated on real-world earthquake, hurricane, and wildfire datasets, our model achieves significant improvements in both predictive accuracy and causal interpretability. It demonstrates strong robustness in low-data regimes and superior cross-hazard generalization capability.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Complex systems with intricate causal dependencies challenge accurate prediction. Effective modeling requires precise physical process representation, integration of interdependent factors, and incorporation of multi-resolution observational data. These systems manifest in both static scenarios with instantaneous causal chains and temporal scenarios with evolving dynamics, complicating modeling efforts. Current methods struggle to simultaneously handle varying resolutions, capture physical relationships, model causal dependencies, and incorporate temporal dynamics, especially with inconsistently sampled data from diverse sources. We introduce Temporal-SVGDM: Score-based Variational Graphical Diffusion Model for Multi-resolution observations. Our framework constructs individual SDEs for each variable at its native resolution, then couples these SDEs through a causal score mechanism where parent nodes inform child nodes' evolution. This enables unified modeling of both immediate causal effects in static scenarios and evolving dependencies in temporal scenarios. In temporal models, state representations are processed through a sequence prediction model to predict future states based on historical patterns and causal relationships. Experiments on real-world datasets demonstrate improved prediction accuracy and causal understanding compared to existing methods, with robust performance under varying levels of background knowledge. Our model exhibits graceful degradation across different disaster types, successfully handling both static earthquake scenarios and temporal hurricane and wildfire scenarios, while maintaining superior performance even with limited data.

Problem

• Modeling complex systems with intricate causal dependencies
• Integrating multi-resolution observational data effectively
• Handling static and temporal scenarios simultaneously

Innovation

• Score-based Variational Graphical Diffusion Model
• Multi-resolution SDEs with causal coupling
• Sequence prediction for temporal dynamics