What we study

Research

Quantile regression and beyond, from theory to real data.

Quantile regression runs through much of our work, but it is only one part of it. We also work on survival analysis, high-dimensional data, and Bayesian models, and we apply them with collaborators in health, aging, and the environment. Most projects start from a real question and end as an R package that others can use.

Quantile Regression

Our core method. We build estimators and tests for quantiles with censored, repeated, and high-dimensional data, at one quantile level or over a whole range of levels.

Survival & Residual Life

Quantiles fit time-to-event data well. We study quantile residual life, which asks how much time is left for someone who has been event-free so far, along with competing risks and patients who may never have the event.

High-Dimensional Statistics

Genomic and metabolomic studies give us thousands of features for a few hundred people. We build quantile-based screening and selection with false discovery rate control, and partial linear models that mix sparse and flexible parts.

Bayesian Modelings

We build Bayesian quantile models for data spread over space and for survival outcomes, with fast variational methods so they scale to thousands of sites.

Machine Learning

We bring machine learning into statistical questions, such as how much of a health disparity an intervention could close, while keeping clear targets and honest uncertainty.

Health & Biomedical Data

We work with epidemiologists and nutrition scientists on cancer, aging, and diet, often with large cohort and trial data from NIH studies.

Get in touch

Questions about our work, a collaboration, or joining the lab? We would like to hear from you.

Email

sangkyulee (at) konkuk.ac.kr

Office

Office 704, Sanghuh Research Building
120 Neungdong-ro, Gwangjin-gu
Seoul 05029, South Korea