Research
Research Themes
Data Science / Machine Learning for Smart and Resilient Critical Infrastructure Systems
We develop novel statistical machine learning approaches, such as Bayesian learning models and neural network models (in particular, graph neural networks), to support the modeling and assessment of urban systems.
Recent publications
Jin-Zhu Yu, Mackenzie Whitman, Amirhassan Kermanshah, and Hiba Baroud. "Hierarchical Bayesian approach for assessing infrastructure networks serviceability under uncertainty." Reliability Engineering & System Safety (2021): 107735.
Jin-Zhu Yu and Hiba Baroud. "Quantifying community resilience using hierarchical Bayesian kernel methods: A case study on recovery from power outages." Risk Analysis 39, no. 9 (2019): 1930-1948.
Jin-Zhu Yu, Mincheng Wu, Gisela Bichler, Felipe Aros-Vera, and Jianxi Gao. (2023). Reconstructing sparse multiplex networks with application to covert networks. Entropy 25, no. 1: 142. (Special Issue Dynamics of Complex Networks)
Jin-Zhu Yu and Hiba Baroud. (2023+). Approximate Gibbs sampler for efficient inference of hierarchical Bayesian models for grouped count data. (Under review at Journal of Statistical Computation and Simulation)
Media
This video was made out of an NSF project titled Bayesian Methods for the Data-Driven Recovery of Networks (2017-2020). The goal was to develop a new Bayesian learning method to improve (i) the ability to model the performance of infrastructure networks given limited data, and (ii) optimize the recovery of these networks following a disruption under uncertainty. Paper No. 2 included above is one of the outcomes of this project.
Decision Analytics for Disaster Management
We develop mathematical programming models, particularly mixed-integer programs and stochastic programs, to aid the decision-making of utility managers in the management of critical infrastructure systems before and after disruptions, such as extreme weather events and malicious attacks.
Recent publications
Jin-Zhu Yu, Wang Yu, Hiba Baroud. Comparing topology-based and flow-based resilience assessment of interdependent infrastructure networks. (2021). Proceedings of the 13th International Conference on Structural Safety and Reliability.
Jin-Zhu Yu, Hiba Baroud, and Haoxiang Yang. (2024+). Two-stage stochastic programming approach for the integrated preparedness and response of interdependent infrastructure networks under uncertainty.
Network Analytics and Resilience
We investigate how the interplay between the dynamic properties or functions of networks, such as resilience, and network structures are impacted by disturbances. For example, how the systemic risk in production networks is affected by demand fluctuations and how the optimal flow network evolves in response to disruptions.
Recent publications
Jin-Zhu Yu, Chencheng Cai, Jianxi Gao, and Jing Wu. (2024+). Bullwhip effect of supply networks: Joint impact of network structure and market demand. (In revision)
Transportation Network Analysis and Control
Transportation networks with human interactions are a highly complex nonlinear dynamical system. We leverage theories and algorithms in network science to make transportation networks smarter, particularly when traffic is disturbed. For example, how to characterize the evolution of congested freeway traffic networks under various recurrent and non-recurrent disturbances.
Recent publications
Yu, J. Z., Wei, Y., He, X., & Gao, J. (2024+). Metastability of congested freeway traffic at road segments. (In preparation)
Other Topics
Community resilience; Climate change adaptation and mitigation; Urban mobility; Decision-making under uncertainty; Bayesian methods