*********************************
There is now a CONTENT FREEZE for Mercury while we switch to a new platform. It began on Friday, March 10 at 6pm and will end on Wednesday, March 15 at noon. No new content can be created during this time, but all material in the system as of the beginning of the freeze will be migrated to the new platform, including users and groups. Functionally the new site is identical to the old one. webteam@gatech.edu
*********************************
Title : Modeling, Learning, Predicting, And Guiding Dynamics Processes Over Networks
Yichen Wang
School of Mathematics
School of Computational Science and Engineering Georgia Institute of Technology
Date : Tuesday, October 03, 2017
Time : 11:00 AM to 1:00 PM EST
Location : KACB 1212
Committee
-------------
Dr. Le Song (Advisor), School of Computational Science and Engineering, Georgia Institute of Technology Dr. Hongyuan Zha, School of Computational Science and Engineering, Georgia Institute of Technology Dr. Haomin Zhou, School of Mathematics, Georgia Institute of Technology Dr. Mark Davenport, School of Electrical and Computer Engineering, Georgia Institute of Technology Dr. Xiaojing Ye, School of Mathematics, Georgia State University
Abstract
-------------
The prevalence of large-scale and fine-grained temporal and spatial event data from dynamic processes over networks poses significant challenges in modeling, learning, predicting and guiding these processes. To tackle these challenges, the thesis aims to develop a unifying framework based on the theory of spatial-temporal point processes, mathematical optimization and stochastic optimal control. In many real world problems, such as modeling the co-evolution of information diffusion and networks, learning time-varying features from user-product purchase graph for recommendation, and promoting the user activities in online discussion forums, the proposed framework leads to novel and elegant models, efficient and scalable algorithms, and the-state-of-art predictive performance.
In this dissertation, we present a novel probabilistic framework for modeling, learning, inferring, and guiding users’ dynamic temporal behaviors. Within the proposed framework, we introduce a pipeline of newly developed statistical models and state-of-the-arts learning algorithms to tackle several canonical problems in practice, including : (1) nonparametric estimation of point process based event patterns, (2) scalable prediction algorithms for point processes, (3) continuous-time closed-loop user activity guiding, and (4) time-sensitive recommendation.
