Online Program Home
My Program

Abstract Details

Activity Number: 127 - SPEED: Statistical Learning and Data Science Speed Session 1, Part 1
Type: Contributed
Date/Time: Monday, July 29, 2019 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistical Learning and Data Science
Abstract #304687 Presentation
Title: Equilibrium Metrics for Dynamic Supply-Demand Networks
Author(s): Fan Zhou* and Hongtu Zhu and Jieping Ye
Companies: University of North Carolina at Chapel Hill and DiDi Chuxing and UNC-Chapel Hill and Didi Chuxing
Keywords: Equilibrium Metrics; Spatio-Temporal Dynamic Systems; Unbalanced Optimal Transport; Functional PCA; Order Dispatching

We introduce a novel class of equilibrium metrics (EMs) to quantify spatio-temporal equilibrium of dynamic supply-demand networks defined on the same graph. It is primarily motivated by measuring the local and global spatio-temporal equilibrium between demand and supply networks in large-scale ride-sharing platforms, such as Uber, Lyft, and DiDi. The two key components of EMs are to formulate the spatio-temporal equilibrium problem as an unbalanced optimal transport problem and to develop an efficient linear programming algorithm to solve such transport problem. Specifically, our EMs measure the local (or global) distance between demand and supply patterns after the optimal transport, while incorporating the related transporting cost. We examine the performance of EM in two important applications, including the use of EMs for predicting local and global answer rates and large-scale order dispatch in ride-sharing platform.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program