Published papers include 'Linear representations of neural networks', 'Adversarial training with general p norms', 'Traffic Forecasting As A Matrix Completion Problem', 'Position-Weighted Backpressure Intersection Control', 'Stochastic Lagrangian Traffic Dynamics', 'Sparse Estimation of Travel Times from Streaming Data', 'Node Modeling for Urban Networks', 'Probabilistic Fundamental Relations', 'Incident Localization and Sensor Placement'.
Research Experience
Research projects involve examining the vulnerability of deep neural networks (DNNs) used in AV control, developing scaling laws to disentangle transient fluctuations from long-term dynamics in traffic, and creating novel operators to solve forward and backward problems in traffic flow.
Background
Research interests include using applied probability, statistical physics, optimization, linear algebra, and dynamical systems theory to address three classes of problems: AI and AV security, the statistical physics of traffic, and neural operators for traffic flow.