Thanks for visiting my site! I am a 5th-year PhD student (2 years for Master's) in statistics at the Oregon State University (OSU). Broadly speaking, my research interests lie in

  • predictive modeling,
  • computational statistics, and
  • machine learning

for large and complex spatial data and time series, with applications to forestry monitoring and energy forecasting. My other interests include deep learning and its more flexible and explainable probabilistic variant models (e.g. deep Gaussian processes, deep Gaussian Markov Random Fields) for space and space-time modeling.

Since Fall 2021, my co-advisors Dr. Lisa Madsen & Dr. Charlotte Wickham (now at Posit) and I have focused on spatial prediction, applied Bayesian, and machine learning for forestry inventory prediction and wind speed forecasting. We focus on Gaussian processes (GPs) and its variant models that are robust, flexible and scalable for larger datasets.

**PhD** My PhD research builds upon the Markov MTD (mixture transition distribution) time series models and Bayesian NNMP (nearest-neighbor mixture process) spatial models developed by Zheng et al. for time series forecasting and spatial prediction in this, this, and this paper. To achieve robust, flexible, and scalable modeling of time series and spatial data, both the MTD and NNMP models utilize probabilistic graphical model (DAG) / Bayesian network representations, finite mixture model structures, Bayesian hierarchical model formations, and MCMC algorithms.

My specific contributions include incorporating copulas in the MTD model to capture more complex and flexible dependence structures, extending the NNMP model to handle zero-inflated skewed data, validating the models and assessing model performance through simulations and applications to real-world data, such as wind energy forecasting and forestry inventory prediction, and compare the models with other benchmark models to ensure robustness and accuracy.

I have also completed all PhD coursework in statistics (e.g. decision theory, probability theory, linear model theory, bayesian statistics). To further improve my algorithm design & model development skillset, I also enroll in several specialization courses in statistics and computer science (e.g. mastering software development in R, data structures and algorithms, deep learning, tensorflow: advanced techniques) on Coursera.

**Master's** For Master's capstone project, advised by Dr. James Molyneux, I worked on selected spatial and temporal point / stochastic processes leading up to Hawkes processes (or self-exciting point processes), which find applications in fields such as seismology, criminology, risk management, finance, etc.

Prior to OSU, I attended Colorado State University (CSU)'s Master of Applied Statistics (MAS) for a semester. I also worked a couple of years in marketing research at two non-profit organizations and graduated with a BS in geosciences from the University of Washington.

When I am not musing on math or stats or programming in R / Python, I enjoy skiing, traveling, being outdoors, and watching thriller and dark comedy films / TV series on Netflix.

Work: [See my projects] Fun: [See my gallery]