I'm a designer of urban, and my tools are data, statistics, and machine learning.

I use data mining and representation learning to understand and design cities. Cities are inherently complex—layered with human behavior, infrastructure, mobility, and social dynamics. To capture this complexity, I work with non-Euclidean structures like graphs and manifolds, because traditional methods fall short when dealing with relational and spatial intricacies. My approach is grounded in unsupervised learning: clustering, community detection, anomaly detection, link prediction, and predictive modeling. These methods help uncover the hidden structures within urban data and translate them into meaningful representations. I also extend this thinking to healthcare analytics, where I've analyzed step count data from the Wrist Doctor 9988 project and am conducting ongoing research—because both domains are fundamentally about understanding people and improving their lives.

For me, a good model isn't just one that works—it's one that makes sense. Interpretability matters because insights, not just predictions, drive better design. Every analysis, every model is ultimately in service of creating cities that are not just efficient, but better and warmer places to live. Ultimately, I'm here to design urban systems through data—to solve real problems and create meaningful change.

• M.S. in Statistical Data Science, University of Seoul
• B.S. in Statistics, Jeonbuk National University

• Developing molecular toxicity link-prediction models based on CTD data
• Developing graph-level anomaly detection models with transformers
• 'Wrist Doctor 9988' Step Count Data Analysis

• Python, R, Julia, SPSS
• Pytorch, Tensorflow
• MySQL, PostgreSQL, duckDB