Zhenhui (Jessie) Li, of Pennsylvania State University, presented Mining Human Mobility Data on April 13 in 502 Davis at 4 p.m.
Abstract
Large-scale human mobility data can be collected from mobile phones, car navigation systems, road surveillance cameras, and location-based applications. Turning such raw data into knowledge can provide insights about our city and empower the city to be more intelligent. In this talk, I would like to share new data mining and machine learning technologies to understand human mobility data. First, I will discuss a new space distance defined by mobility flow. We propose region representation learning via mobility flow and demonstrate the use of such representations in predicting crimes and region properties. Next, I will introduce our novel spatial-temporal deep learning models that demonstrate superior performance in predicting taxi demands. The model is validated using real large-scale Didi Chuxing data. Lastly, I would like to share an opportunity of City Brain project, which I will soon start working on from summer 2018. This project will be in collaboration with traffic police department of Hangzhou, China. We will have access to real-time traffic data from thousands of road surveillance cameras in the city and we will conduct field experiments to actually control the traffic signals in the city. We will be exploring AI-empowered data-driven traffic signal control system. I am looking for potential collaborations on this exciting project.