Tianjiao Li

About Me

I am a fifth-and-final-year PhD student in Operations Research at ISyE, Georgia Tech. I am fortunate to be advised by Prof. Guanghui (George) Lan and Prof. Ashwin Pananjady. I also obtained an M.S. in Quantitative and Computational Finance from the same institute. Before that, I obtained my B.S. in Mathematics from Fudan University.

I will be a Postdoctoral Researcher at the MIT Sloan School of Management, starting in Fall 2025, working with Prof. Swati Gupta.

Research Interests

My research focuses on the design and analysis of efficient algorithms for Nonlinear Optimization, Stochastic Optimization, and Dynamic Decision-Making. I am particularly interested in

• Parameter-free methods for convex and nonconvex optimization
• Stochastic optimization for statistical learning and machine learning
• Policy optimization and policy evaluation for Markov decision processes
• Applications: healthcare, E-commerce, finance, etc

Publications

Preprints

Awards

• Alice and John Jarvis Best Student Paper Award, 2024
  • Awarded annually to one Ph.D. student in ISyE across all disciplines
• Shabbir Ahmed PhD Fellowship for Excellence in Research, 2023
  • Awarded annually to one Ph.D. student in ISyE for research in optimization

• Second Place, Poster Competition, YinzOR Student Conference 2024

• First Place, Best Poster Award, Georgia Statistics Day 2023

Teaching

• Course Instructor, Summer 2024, Georgia Tech
  • Course: Statistics and Applications (ISyE 3770)
  • Description: Introductory probability and statistics course for engineering and computer science students
  • Class size: 64 (26 on campus + 38 online)
  • Teaching evaluation: 4.8/5.0 (response rate: 56%)
    • Respect for students: 4.8/5.0
    • Inclusiveness: 4.9/5.0
    • Communicated how to succeed: 4.7/5.0
    • Availability: 4.9/5.0
    • Stimulates interest: 4.6/5.0
    • Clarity: 4.5/5.0
    • Feedback helpfulness: 4.8/5.0
• Guest Lecturer, Fall 2024, Georgia Tech
  • Course: Computational Data Analysis / Machine Learning (ISyE 6740)
  • Instructor: Guanghui (George) Lan
  • Responsibility: 2 Lectures in machine learning and data analysis
• Guest Lecturer, Spring 2024, Georgia Tech
  • Course: Optimization Methods for Reinforcement Learning (ISyE 8803)
  • Instructor: Guanghui (George) Lan
  • Responsibility: 8 Lectures in optimization methods for policy evaluation and average-reward MDPs

Talks and Presentations

• ORIE Colloquium, Cornell University, Ithaca, NY, Feb 2025
  • Title: Universal Parameter-Free Methods for Convex, Nonconvex, and Stochastic Optimization
• STOR Colloquium, UNC Chapel Hill, Chapel Hill, NC, Jan 2025
  • Title: Universal Parameter-Free Methods for Convex, Nonconvex, and Stochastic Optimization
• INFORMS Annual Meeting, Seattle, WA, Oct 2024
  • Session: First-Order Methods in Continuous and Stochastic Optimization
  • Title: A Simple Uniformly Optimal Method without Line Search for Convex Optimization
• Cornell ORIE Young Researchers Workshop, Ithaca, NY, Oct 2024
  • Session: Optimization I
  • Title: A Simple Uniformly Optimal Method without Line Search for Convex Optimization
• YinzOR Student Conference, CMU Tepper School of Business, Pittsburg, PA, Aug 2024
  • Poster Presentation: Accelerated Stochastic Approximation with State-Dependent Noise
  • Won the Second Place in the poster competition
• International Symposium on Mathematical Programming (ISMP 2024), Montreal, Canada, Jul 2024
  • Session: Advances in Stochastic First-Order Methods
  • Title: A Simple Uniformly Optimal Method without Line Search for Convex Optimization
• DAO Team Seminar at Laboratoire Jean Kuntzmann, Université Grenoble Alpes, Grenoble, France, May 2024
  • Title: A Simple Uniformly Optimal Method without Line Search for Convex Optimization
• INFORMS Optimization Society Conference, Houston, TX, Mar 2024
  • Session: Advances in Continuous Optimization Algorithms
  • Title: A Simple Uniformly Optimal Method without Line Search for Convex Optimization
• INFORMS Annual Meeting, Phoenix, AZ, Oct 2023
  • Session: Recent Advances in Policy Optimization and Reinforcement Learning
  • Title: Accelerated and Instance-Optimal Policy Evaluation with Linear Function Approximation
• Georgia Statistics Day 2023, Atlanta, GA, Oct 2023
  • Poster Presentation: Accelerated and Instance-Optimal Policy Evaluation with Linear Function Approximation
  • Won the First Place in the poster competition
• SIAM Conference on Optimization, Seattle, WA, May 2023
  • Session: New Sparse Optimization
  • Title: Accelerated Stochastic Approximation with State-Dependent Noise
• INFORMS Annual Meeting, Indianapolis, IN, Oct 2022
  • Session: Reinforcement Learning Theory
  • Title: Stochastic First-Order Methods for Average-Reward Markov Decision Processes
• ISyE Ph.D. Student Seminar, Atlanta, GA, Sept 2022
  • Title: Stochastic First-Order Methods for Average-Reward Markov Decision Processes
• Asilomar Conference on Signals, Systems, and Computers, Online, Nov 2021
  • Session: Theory of Reinforcement Learning
  • Title: Faster Algorithm and Sharper Analysis for Constrained Markov Decision Process
• INFORMS Annual Meeting, Online, Oct 2021
  • Session: Stochastic Optimization in Machine Learning
  • Title: Simple and Optimal Methods for Stochastic Variational Inequalities

Services

Journal Reviews

• SIAM Journal on Optimization
• Mathematical Programming
• Annuals of Statistics
• Computational Optimization and Applications
• Optimization Letters
• Journal of Global Optimization

Conference Reviews

• Annual Conference on Learning Theory (COLT) 2022-2025

Session Organization

• INFORMS Annual Meeting 2024, Seattle, WA, Oct 2024
• Session: Advances in Continuous Optimization Algorithms
• Session: Advances in Non-Smooth Optimization
• International Symposium on Mathematical Programming (ISMP 2024), Montreal, Canada, Jul 2024
• Session: Advances in First-Order Methods for Continuous and Stochastic Optimization

Powered by Jekyll and Minimal Light theme.