Projects & publications

The emergent complexity of contagion

Spreading patterns that look like complex contagion can sometimes arise from simpler contagion rules that vary across a network and are correlated with community structure. I work on statistical methods to detect and measure this kind of emergent complexity in contagion data and to understand when an underlying complex contagion mechanism can be distinguished from hidden heterogeneity. More broadly, I am interested in when observed spreading patterns provide reliable evidence about how contagion actually works.

Hypergraph reconstruction from dynamics

Many complex systems involve interactions among groups of components, but recovering this structure from noisy dynamical data is difficult. This project develops a Bayesian extension of sparse, Taylor-based hypergraph inference that quantifies uncertainty and helps determine which higher-order interactions are supported by data. More broadly, I am interested in understanding when higher-order structure is fundamentally identifiable from observations of dynamics.

Pressure points in decarbonized power grids

Reliability failures in decarbonized power grids with renewable energies often emerge from interactions across generation, transmission, storage, and demand rather than from a single bottleneck. This project uses transfer entropy to infer predictive pathways of stress propagation through simulated power systems. The goal is to identify nonlocal, scenario-dependent reliability risks that may not be apparent from the physical network or from individual component bottlenecks alone.