Modeling, predicting, and guiding users' temporal behaviors

The increasing availability and granularity of temporal event data produced from user activities in online media, social networks and health informatics provide new opportunities and challenges to model and understand user behaviors. In addition to studying the macroscopic patterns on the population level, such type of data further enable us to investigate user interactions in a more fine-grained scale to address the "who will do what by when?" question with new exploratory and predictive models. On the other hand, these myriads of microscopic event data, such as publishing a post, forwarding a tweet, purchasing a product, checking in a place, often arise asynchronously and interdependently; hence they require new representing and analyzing methods far beyond those based on independent and identically distributed data models. In this dissertation, I present a novel probabilistic framework for modeling, learning, predicting, and guiding users’ temporal behaviors. Within the proposed framework, we introduce a pipeline of newly developed statistical models, state-of-the-arts learning algorithms to tackle several canonical problems in theory and practice, including: (1) provable nonparametric learning of temporal point processes, (2) a generic embedding framework for continuous-time evolving graphs, (3) scalable algorithms for predicting user activity levels, and (4) a stochastic differential equation framework for guiding users’ activities.

Level thesis:degree_level

Doctoral

Department dc:contributor.department

Mathematics

Grantor dc:publisher

Georgia Institute of Technology

Year dc:date.issued

2018

Author dc:creator

• Wang, Yichen

Advisor dc:contributor.advisor

• Song, Le

Committee members dc:contributor.committeemember

• Zha, Hongyuan
• Davenport, Mark
• Ye, Xiaojing
• Zhou, Haomin

• Point processes
• Hawkes processes
• Survival analysis
• Low-rank models
• Mass transport
• Fokker Planck equation
• Stochastic optimal control
• Reinforcement learning
• Social network analysis
• Information diffusion
• Recommendation systems

Language dc:language.iso

en_US

Handle dc:identifier.uri

http://hdl.handle.net/1853/60208

OAI identifier oai:identifier

oai:repository.gatech.edu:1853/60208