BIOS Seminar: Statistical Methods in Ride-sharing Platforms

Hongtu Zhu, PhD, professor in the Department of Biostatistics, will present this “Modern Statistical Methods in Ride-sharing Platforms and other IT companies.”

Dr. Hongtu Zhu, has a broad background in statistics, biostatistics, medical imaging, genetics and computational neuroscience, with specific training and expertise in neuroimaging data analysis and big data integration as well as secondary data analysis on neurodegenerative and neuropsychiatric diseases. As a graduate student and postdoctoral fellow, he developed various new statistical methods for analyzing genetic, behavioral and clinical data from cross-sectional, longitudinal and family studies, and solved associated statistical issues (e.g., estimation, hypothesis testing). As a faculty member at Columbia University and New York State Psychiatric Institute, he expanded his research to develop new statistical methods for the analysis of medical imaging data including magnetic resonance images (MRI), DTI, EEG/MEG, Ultra Sound, fMRI and PET.

Abstract:

Two fundamental questions of interest for ride-sharing platforms (e.g., Uber and DiDi) are how to evaluate whether a ride-sharing platform achieves a healthy equilibrium between
dynamic supply and demand systems, called a market equilibrium, or not and how to achieve such healthy equilibrium. Solving these questions require big advancement
in various analytical methods.The first aim of this talk is to illustrate how some existing statistical methods (e.g., statistical methods for recurrent events）can be used to analyze big customer data in order to improve the management of ridesharing platforms. The second aim of this talk is to illustrate why other data analytical methods (e.g., xgboost, deep
learning, Markov decision process, Graph network data analysis, matching methods) are critical for solving many other analytical tasks (e.g., pricing, dispatching, mapping, ride
sharing,recommendation system, face recognition, intelligence customer service). Some real data analysis results are used to illustrate these methods. Most of these methods can be used to promote the efficiency of medical practices.