Haoming Yang

Ph.D

Engineering Technology

Future Mobility Systems Lab Carnegie Mellon University

PhD Research Assistant

- Skills: Graph Neural Networks, Attention Mechanism, Agent-based Simulation, Web Crawling, AWS - Developed ETL pipeline in Python to aggregate multiple types of datasets (e.g., demographic data, transit schedule, street maps, travel surveys) for large-scale, agent-based transportation simulation - Configured and deployed simulation environment on AWS, executing scenarios with shared autonomous vehicles (SAVs) in network consisting of 27,000 nodes and 57,000 links - Conducted sensitivity analysis on pricing and fleet size, evaluated impacts of SAVs on ride sharing, equity and network - Proposed a novel algorithm to impute missing spatiotemporal data, developed a multi-head attention mechanism that maps - Incorporated attention mechanism to graph-based spatiotemporal convolutional networks for imputation and inference