Prof. Changhee Lee

About Me

I am currently an assistant professor at the Department of Artificial Intelligence, Chung-Ang University, Seoul, Korea. I am very excited and motivated by the promise of revolutionalizing domains -- especially healthcare and manufacturing -- where critical decision-making is the key through the use of machine learning in combination with domain experts.

In particular, my research interest lies in integrating multiple modalities, building individualized models of care via prognostic and causal inference models, discovering scientific knowledge from data, interpreting "black-box" machine learning methods, and applying advanced machine learning methods at scale.

I completed my Ph.D. at the University of California, Los Angeles, Department of Electrical and Computer Engineering, as a member of the van der Schaar Lab led by my advisor Prof. Mihaela van der Schaar. My research focus was on deep learning approaches to address challenges associated with modeling, predicting, and interpreting in time-to-event analysis and time-series analysis and on deep learning approaches for multiple omics data including genomics and transcriptomics.

Before joining the van der Schaar Lab, I completed my B.S. and M.S. in Electrical and Computer Engineering at Korea University in 2011 and 2013, respectively. I then worked as a research staff in Electronics and Telecommunications Research Institute (ETRI) developing methods for next-generation user-centric wireless communications and being involved in 3GPP RAN1/RAN2 standardization activities.

Education

University of California, Los Angeles, USA (Aug. 2016 - Aug. 2021)


Korea University, South Korea (Mar. 2011 - Feb. 2013)

  • M.S. in Electrical and Computer Engineering

  • Advisor: Prof. Inkyu Lee (이인규 교수)


Korea University, South Korea (Mar. 2007 - Feb. 2011)

  • B.S. in Electronic and Electrical Engineering

Research Experience

van der Schaar Lab, UCLA, United States (Sep. 2016 - Aug. 2021)

Graduate Student Researcher

  • Designed deep learning approaches for both static and longitudinal time-to-event analyses under competing risks.

  • Proposed outcome-oriented temporal clustering methods to discover patient subgroups with similar outcomes based on clinical trajectories.

  • Developed an information bottleneck approach for integrating incomplete multi-omics data with various omics-missing patterns.

  • Designed a deep learning approach to adjust for various covariate shifts in heterogeneous treatment effect estimation based on observational time-to-event data.

  • Proposed a self-supervised learning framework for feature selection under the low labeled data regime.


Public Health England, United Kingdom (Mar. 2019 – Apr. 2019)

Visiting Researcher

  • Developed a novel temporal phenotyping method on stage III breast cancer cohorts in the UK.


Electronics and Telecommunications Research Institute (ETRI), South Korea (Mar. 2013 – May 2016)

Research staff

  • Developed next-generation wireless communication methods for user-centric 5G Mobile Personal Cells (MPC).

  • Involved in 3GPP RAN1 standardization activities on “LTE Device-to-Device Proximity Services” and RAN1/RAN2 standardization activities on “LTE Narrowband Internet-of-Things”.


Wireless Communications Lab, Korea University, South Korea (Mar. 2011 – Feb. 2013)

Graduate Student Researcher

  • Developed optimal antenna location designs for distributed antenna systems and transmission schemes for inter-cell interference cancellation in multi-cell systems.