Robbert van der Burg
PhD candidate
VU University Amsterdam
I am a first-year PhD candidate in the Department of Mathematics at the Vrije Universiteit Amsterdam. My research lies at the intersection of probability theory, network science, and applied statistics, under the supervision of Alessandro Zocca and Frank van der Meulen.
My PhD research focuses on probabilistic modeling of weather-driven failures in critical infrastructure networks, such as power grids, as part of the VIDI "Power Network Optimization in the Age of Climate Extremes" (https://doi.org/10.61686/GOOEL09973). These networks are inherently spatially distributed, and when subjected to extreme environmental stressors, such as wind storms, floods, or heatwaves, failures rarely occur in isolation. Instead, extreme weather manifests as a spatial field with a characteristic footprint, simultaneously affecting multiple components of the network. As a result, the vulnerability of nodes is strongly shaped by spatial dependence and correlation structures.
In addition to my main PhD topic, I maintain a strong interest in Markov chain theory and network analysis. Related side projects include the study of hitting times on graph models, first passage times, and the Kemeny constant, as well as applications of these concepts in social network analysis. I frequently work with Bernd Heidergott and Thao Le on projects spanning random walks on networks, opinion dynamics, and infrastructure reliability.
Research Interests
Random Walks on Networks
Non-standard random walks with edge weights, optimization problems, and connections to Kemeny constants. Collaborators: Thao Le, Alessandro Zocca, Bernd Heidergott and Ines Lindner.
Infrastructure Failure Modeling
Comparing probabilistic models (Ising, Bayesian hierarchical, independent Bernoulli, autologistic regression) for analyzing cascading failures in infrastructure networks.
Opinion Dynamics & QSD Theory
Connecting quasi-stationary distribution theory to opinion dynamics models (DeGroot, Friedkin-Johnsen), extending BSc thesis work on metastable behavior.
Latest News
NetSci Presentation! I presented recent joint work with Alessandro Zocca and Frank van der Meulen at the NetSci Summer Symposium 2026. The talk, Bayesian Recovery of Dependence Structures on Networks, explores how empirical Bayesian parameter estimation can be used to distinguish genuine network-driven dependence from correlations induced by external covariates. The project combines covariate-dependent Ising models with spike-and-slab priors to recover dependence structures while quantifying uncertainty in the inferred network. A preprint describing the methodology and results will be available soon. Slides here.
New Preprint available! Together with Thao Le, Bernd Heidergott, Ines Lindner and Alessandro Zocca, we have just finished a new paper titled 'Random Walks with Traversal Costs: Variance-Aware Performance Analysis and Network Optimization'. We introduce weighted Markovian graphs, a framework that decouples random walk dynamics from (possibly stochastic) edge traversal costs, and derive closed-form expressions for the mean and variance of weighted first passage times and Kemeny constants. We showcase the framework through two applications to surveillance patrol policies and traffic network optimization under disruptions. Preprint available on arXiv.
Phd Started! Began PhD in Mathematics at VU University Amsterdam, working on probability theory, network science, and applied statistics.