Quan Guo

Quan Guo

Postdoc at LANL

Quan Guo

About me:

I am a Postdoc in Los Alamos National Lab (LANL). Before coming to LANL, I studied in Georgia Institute of Technology (Georgia Tech) where I obtained a Ph.D of Water Resource Engineering (advised by Prof. Jian Luo) and Masters of Environmental Engineering and Computational Science & Engineering.

What kind of person I am:

  • Passionate cross-disciplinary researcher in Physics and Machine Learning, dedicated to utilizing AI and machine learning to solve scientific problems and construct digital twins.
  • Python Research and Development (R&D) since 2018, and deeply committed to continuous learning for efficient continuous integration and continuous deployment (CI/CD).
  • Combining traditional domain knowledge and statistical analysis with modern big data and AI to make robust, scaling and trustworthy predictive models.

Don’t hesitate to contact me:

  • Email: qguo48@hotmail.com

Research interests:

My research is at the intersection of data science, machine learning, and traditional physics. Specifically, I employ physics-informed deep learning and scientific AI techniques to tackle geophysics inverse problems, particularly those related to groundwater. Here are some methods I am using:

  • Physics-Informed Neural Networks: Combining physics-based knowledge to train neural networks with sparse data while enhancing generality and robustness.

  • Neural Operators: Exploring the use of neural operators as surrogate forward models to efficiently solve high-dimensional PDE and simulate physical process.

  • Generative Modeling: Implementing AI generative models like MoCoGAN, StyleGAN, Normalizing Flow, etc to recognize and encode content and motion of images and videos, specifically for subsurface realizations.

  • Bayesian Analysis and High-Performance Computing: Employing Bayesian analysis and high-performance computing to expedite numerical inverse optimization and uncertainty quantification.

  • Big data and Cloud Computing: Developing Spark and Hadoop softwares for multi-source data assimilation and geospatial feature engineering