Spectral graph theory
Using spectral perspectives to understand graph structure, operators, and learning behavior.
Research
I study graph-structured intelligence for dynamic, uncertain, and interconnected systems, with current emphasis on theory, cross-network dynamics, and AI systems.
Current Directions
UQ for graph dynamics
Studying uncertainty quantification, inference, and intervention in graph-driven dynamical systems.
Cross-network graph dynamics
Understanding how heterogeneous networks interact, couple, and evolve together over time.
General theory for graph dynamics
Searching for broader principles that unify dynamics, interaction, and controllability on graphs.
Selected Work
Unified GNN framework
Survey and tutorial-building effort toward unifying spectral and spatial graph learning.
Influence on graphs
Graph Bayesian Optimization, source localization, and influence maximization for dynamic and multiplex networks.
Spatiotemporal and urban forecasting
Graph-temporal forecasting for traffic and mobility, including award-winning work on urban intelligence.
Materials and scientific discovery
Graph/ML-enabled discovery for solid-state lithium-ion conductors and related science-facing systems.