Dahyun Jeong

Integrated Master's & Doctoral Program
Department of Statistics, University of Seoul

Google Scholar
Email: dahyunjeong25@gmail.com

I am currently enrolled in the Integrated Master's & Doctoral Program at the University of Seoul, advised by Prof. Yunjin Choi. I am interested in methodologies for discovering latent structures in complex data and modeling them statistically. My current research centers on clustering and topic modeling, with a particular focus on modeling structural relationships when multiple sources of data are observed simultaneously. I also aim to discover new research questions from real-world applications, formulate them mathematically, and develop them into statistical learning and optimization methodologies.

Research Interests

Statistical learning
I study methodologies for discovering and modeling latent structures in data, with a focus on unsupervised learning. My main interests include clustering, subspace learning, and topic modeling.

Modeling data with complex structures
For data collected from multiple sources or from multiple related problems, I focus on methods that identify and leverage both the features shared across them and the features distinct to each.

Optimization
I am interested in convex and nonconvex optimization methods for statistical learning problems, and in the development and application of optimization algorithms such as ADMM.

Publications

Beyond Local Centers toward Prototypes: Joint Learning of Self-Representation and Adaptive Clustering
D. Jeong, Y. Choi
ACM International Conference on Information and Knowledge Management (CIKM) (2026, to appear)

Automatic Seed Word Selection for Topic Modeling
D. Jeong, J. Hwang, Y. Choi, Y. Y. Kim
IEEE Access, 13, 31269–31285 (2025)

Enhancing Social Media Post Popularity Prediction with Visual Content
D. Jeong, H. Son, Y. Choi, K. Kim
Journal of the Korean Statistical Society, 53(3), 844–882 (2024)

Determinants of Smart Healthcare Utilization: A Focus on Individual Characteristics and Goal-setting Behavior
D. Jeong, H. Son
Journal of Health Informatics and Statistics, 49(4), 392–403 (2024)

Ongoing Research

Multi-view clustering (to be submitted)
A multi-view clustering method that jointly leverages consensus structure and view-specific structures

Multi-task subspace clustering
A multi-task clustering method that shares information across tasks while preserving the subspace structure of each task

Topic discovery
Topic modeling that separately discovers background and semantic topics

Education

Research Interests

Publications

Ongoing Research

Future Research Interests