Who: Dr. Ran Sun
When: Friday, December 13, 2024, 11:00am-11:50am
Where: Kiewit Hall (KH) Room A510, (Lincoln) Peter Kiewit Institute (PKI) Room 160, (Omaha, remote)
Zoom: https://unl.zoom.us/j/98630335322
Title: Addressing Data Challenges in Transportation Networks: Stochastic Optimization and Machine Learning Approaches
Presentation Abstract: The inherently underdetermined and uncertain nature of complex systems, such as transportation networks, creates barriers to accurate modeling, reliable prediction, and efficient management. Direct data observations from modern transportation and mobility systems often face challenges caused by high-dimensionality, indirect measurements, missing values, and redundancy, which hinder their immediate usability. In this presentation, I will introduce novel stochastic optimization and machine learning methods designed to tackle these data challenges from both demand and supply perspectives in transportation systems. I will demonstrate the developed estimation and prediction methods in network flow, travel demand, logistics, and autonomous driving perception applications. Lastly, I will discuss the potential extensions to broader cases in ITS, safety, and electrification.