General Information

Full Name Hyemin Gu
Languages English, Korean


  • 2020 - Now
    PhD student in Mathematics
    University of Massachusetts Amherst, MA, US
    • Research assistant
    • Specializing in Mathematical modeling
    • Advised by Markos Katsoulakis
    • Current GSAC member - TWIGS coordinator
  • 2018 - 2020
    MSc in Mathematics
    Ewha Womans University, Seoul, Republic of Korea
    • Thesis title "Convolutional Neural Network for 2D Flow Estimation Problem"
    • Specializing in Numerical analysis
    • Advised by June-Yub Lee
  • 2014 - 2018
    BSc in Mathematics
    Ewha Womans University, Seoul, Republic of Korea
    • Major in Mathematics and Computational Science
    • Minor in Statistics


  • 2022 - Now
    Graduate research assistant
    University of Massachusetts - Amherst, Amherst, MA, US
    • Developed a particle transportation algorithm and implemented the algorithm as a generative model.
    • Improved the base model by autoencoders and and proved a sufficient condition for the convergence of the improved model.
  • 2021
    Graduate teaching assistant
    University of Massachusetts - Amherst, Amherst, MA, US
    • TA for M545 (Linear Algebra for Applied Math), M235 (Linear Algebra), M532H (Nonlinear Dynamics & Chaos with Applications).
    • Hosted discussion sessions, delivered mini-courses for Python ODE solving tutorials, and graded weekly assignments.
  • 2020
    Post-master's researcher
    Ewha Womans University Seoul Hospital, Seoul, Korea
    • Constructed a pipeline for gene expression data analysis using R and created a documentation.
    • Trained medical school graduate students to conduct statistical analysis using R.
  • 2018 - 2020
    Graduate research assistant
    Ewha Womans University, Seoul, Korea
    • Trained CNN to estimate a physical state variable from 2D flow velocity fields.


  • 2022 - Now
    Lipschitz regularized generative particles algorithm
    • Implementation of a particle transportation algorithm through gradient flows associated with Lipschitz regularized f-divergences.
    • A generative model alternative to GANs in low training data regimes.
    • Mathematical interpretation of applying spectral normalization on neural network discriminators as a particle transportation speed regularization.

Academic Interests

  • PDE based machine learning
    • Optimal transport, gradient flows, transport equation.
    • Reaction-diffusion equations.
    • Generative models

Other Interests

  • Hobbies: Raising plants at home, visiting art gallaries, learning investment, yoga