Michael Chung, Ph.D.
Researcher, National Center for Toxicological Research, FDA, Jun. 2013 – Dec. 2015
PhD., Mathematics, University of Arkansas, Fayetteville, AR, Dec. 2015
M.S., Statistics, University of Arkansas, Fayetteville, AR, May 2009
B.S., Mathematics, National Central University, Jhongli, Taiwan, Jul. 2004
Before coming to the Brain Imaging Research Center, I accepted a predoctoral fellowship at National Center for Toxicological Research (NCTR), U.S. Food and Drug Administration (FDA). In NCTR, I first designed and developed a specific probabilistic topic models (i.e., a type of latent variable model) to discover hidden biological pathways in large-scale microarray data. Later, I built database from Veterans Affairs electronic medical record through VINCI, which includes over 485 million health factor records. I also helped developed fit-to-purpose probabilistic topic model for VA dataset. I also developed Bayesian nonparametric model to discover multiple social ties in social networks including a gathering of over 14 million edges from Twitter. Through these experiences, I have accumulated knowledge of analyzing big data with high dimensions and complex structure.
After I joined Brain Imaging Research Center (BIRC), I aim to bring my big data modeling experience into the neuroimaging science — specifically, methodologies like deep learning and Bayesian nonparametric models. Under the mentorship of Dr. Clint Kilts (Director of BIRC) and Dr. Keith Bush, I plan to bridge the gap between new machine learning tools and neuroimaging science.
Due to the recent advances of machine learning technologies (e.g., deep learning), there have been many successful applications of using artificial intelligence to solve very complex problems, like computer vision or self-driving cars. Big company like Google, Facebook, and other tech companies are constantly pouring hug resources into improving and expanding machine learning technologies, which often are open-sourced and free to the general public. With my mathematical background and big data experience, I am interested in utilizing the new machine learning tools to overcome challenges in computational neuroscience. My research interests include machine learning, deep learning, and Bayesian nonparametric statistics.
- Chung, M. H., Chen, G., Zhao, W., Hao, G., Pan, J., & Xu, X. (2016). Discovering Multiple Social Ties for Individuals in Online Social Networks. The Third European Network Intelligence Conference. (Accepted)
- Chung, M. H., Wang, Y., Tang, H., Zou, W., Basinger, J., Xu, X., & Tong, W. (2015). Asymmetric author-topic model for knowledge discovering of big data in toxicogenomics.Frontiers in Pharmacology, 6, 81.
- Song, J. J., Yu, P., Ren, Y., & Chung, M. H. (2009). Correspondence analysis for studying association between geography and cancer.Journal of the Korean Data and Information Science Society, 20(5), 919-924.