• Conducting research on mediation analysis with applications to genomic and proteomic data under Prof. Peng Wei at MD Anderson Cancer Center, leading simulation studies, real-data analyses, and manuscript preparation.

• Advancing research in causal inference and machine learning by imputing protein data and developing summary statistics to better estimate causal pathways linking risk factors to health outcomes.

• Supervised by Prof. Chongzhi Di to provide statistical support for the Women’s Health Initiative’s Objective Physical Activity and Cardiovascular Health (OPACH) Study. Responsibilities included processing and analyzing high-resolution data collected from wearable accelerometers and electrocardiogram (ECG) monitors using R.

• Developed an R package for comprehensive accelerometry data analysis, encompassing data preprocessing, visualization, functional principal component analysis, and functional regression for various data types (e.g., continuous, discrete, time-to-event).

• Led a manuscript (as the first author) to adapt functional regression models to investigate dose-response relationships between physical activity accumulation patterns and cardiometabolic risk factors among older women in OPACH. (Refer to the Publications/Preprints section.) 

• Collaborated with psychiatrists to design a clinical trial validating the effectiveness of interactive digital therapeutic games in promoting physical activity to alleviate depressive symptoms. Prepared statistical analyses for IRB submissions, including sample size estimation and hypothesis testing.

• Collected and analyzed longitudinal psychological assessments, physical measurements, HRV, and Kinect motion sensor-based activity vision data from control group participants (N=30) engaged in digital therapeutic games as part of a physical activity research team. Created decision-support documents for the director, including statistical analysis and visualizations using R.

• Developed a manuscript (as the first author) analyzing research trends in fitness and digital health by using natural language processing techniques (Latent Dirichlet Allocation Topic Modeling) to identify key themes across 15,950 papers from Web of Science SCIE journals with R and Python. (now on arXiv)

• Completed mandatory military service as a teaching assistant and caregiver at Seoul Gyeongun School, a special education institution in South Korea, supporting 12 middle school students with severe autism and intellectual disabilities for 2 years.

• Conducted neuroimaging research under Prof. KC Gary Chan to analyze atrophy severity patterns in Alzheimer’s patients using MRI data from the ADNI database collaborating with Profs. David Haynor and Dean Shibata from the UW Department of Radiology. Applied general and partial correlation models, including Gaussian Copula Graphical Models, on multiple distinct localized components derived from Orthogonal Projective Non-negative Matrix Factorization in R. (Published in the Journal of Alzheimer’s Disease)

• Led a manuscript (as the first author) on estimating Alzheimer’s disease progression score, developing a machine learning model (Supervised Variational Autoencoders) in TensorFlow, adapted to leverage baseline neuroimaging features (e.g., gray matter volumes) and demographic data from the ADNI dataset. (First-authored abstracts published at ISMRM and AAIC, with a manuscript currently under journal review)

• Supervised under Prof. Chongzhi Di to focus on research topics such as mobile health, wearable devices, and physical activity. Analyzed accelerometer-measured physical activity accumulation patterns in relation to cardiometabolic health in older women using R.

• Designed and implemented an analytic framework combining behavior recognition (using published machine learning algorithms), feature extraction and functional regression models, and applied the procedure to the OPACH study data (N=6500, Data recorded every 15 seconds over 7 consecutive days per individual).

• Conducted research with Prof. Ali Shojaie (UW Biostatistics) focused on classifying Alzheimer’s disease using fMRI data from the ADNI database.

• Reviewed literature and adapted machine learning/graphical models, including supervised variational autoencoders and omnibus graph embeddings, using R and Python to suit our specific data settings

• Kwanjeong Educational Foundation (KEF) is emerging as the largest educational foundation in Asia, with an estimated value of 1500 billion Korean won (1.15 billion US dollars). 

• Full tuition and fees for 2 years of graduate studies

R, Rmd, Python, Latex, MATLAB, MySQL, Unity

Tableau, E-Views, MS Excel, MS Word, MS Powerpoint

English, Korean, Spanish