CV
Education, research collaborations, publications and awards.
Contact Information
| Name | Jihyeon Seong |
| Professional Title | Ph.D. Student in Artificial Intelligence, KAIST |
| jihyeon.seong@kaist.ac.kr |
Professional Summary
Ph.D. student at the Kim Jaechul Graduate School of AI, KAIST, advised by Prof. Jaesik Choi. Works on uncertainty quantification under distribution shift, time-series representation learning under change points, and explainable AI for clinical decision support.
Experience
-
2023 - Present Research Collaboration — Prediction of Acute Kidney Injury
Seoul National University Bundang Hospital
- Built the residual-CNN prediction model and its layer-wise relevance propagation (LRP) explanations; in a prospective reader study with 11 clinicians, model assistance raised recall from 61.0% to 74.0% and cut mean review time by 8.4 seconds.
-
2024 - 2025 Teaching Assistant
KAIST, Kim Jaechul Graduate School of AI
- Interpretability and Interactivity in AI (Prof. Jaesik Choi) — Fall 2025
- Deep Learning (Prof. Jaegul Choo) — Spring 2025
- Deep Learning (Prof. Jaesik Choi) — Spring 2024
-
2022 - 2022 Research Collaboration — Anomaly Detection in Semiconductor Fabrication
Samsung Device Solutions
- Detected and explained contamination events in a semiconductor fab from multivariate time-series sensor data.
-
2022 - 2022 Research Collaboration — Correction of Electronic Pen Signals
HiDeep
- Designed an autoencoder-based method to correct distortion in electronic pen input signals.
-
2021 - 2021 Research Collaboration — Deep Air, Blast Furnace Permeability Prediction
POSCO Gwangyang Works
- Designed an explainable CNN-based blast furnace permeability prediction model from operational sensor time series.
Education
-
2023 - Present Republic of Korea
Ph.D., Kim Jaechul Graduate School of AI
Korea Advanced Institute of Science and Technology (KAIST)
Artificial Intelligence
- Advisor: Prof. Jaesik Choi
- Expected: Aug 2028 or Feb 2029
-
2021 - 2023 Republic of Korea
M.S., Kim Jaechul Graduate School of AI
Korea Advanced Institute of Science and Technology (KAIST)
Artificial Intelligence
- Advisor: Prof. Jaesik Choi
-
2016 - 2021 Seoul, Republic of Korea
B.S.
Ewha Womans University
Computer Engineering
-
2017 - 2017 Paris, France
CS Global Field Training, summer exchange student
EPITA
Computer Science
Awards
-
2024 Student Travel Grant
International Joint Conference on Artificial Intelligence (IJCAI)
-
2023 Best Paper Award
Korea Computer Congress (KCC)
Capsule Neural Networks as Noise Stabilizer for Time Series Data
-
2022 Digital Idea Market Award (4th place)
Kookmin Bank — ESG AI Investment
-
2020 Merit Scholarship for Academic Excellence (ranked 1st in class)
Ewha Womans University
Publications
-
2026 DistMatch: Adaptive Binning via Distribution Matching for Robust Sequential Conformal Prediction
International Conference on Machine Learning (ICML)
-
2026 Towards Transparent Time Series Analysis: Exploring Methods and Enhancing Interpretability
ACM Computing Surveys
Role: wrote “Limitation of XAI in Time Series”.
-
2025 Validation of an Acute Kidney Injury Prediction Model as a Clinical Decision Support System
Kidney Research and Clinical Practice
Role: model development, software, data curation, methodology.
-
2024 Towards Dynamic Trend Filtering through Trend Point Detection with Reinforcement Learning
International Joint Conference on Artificial Intelligence (IJCAI), pp. 2324–2332
-
2024 Towards Diverse Perspective Learning with Selection over Multiple Temporal Poolings
AAAI Conference on Artificial Intelligence (AAAI)
-
2024 CardioCaps: Attention-Based Capsule Network for Class-Imbalanced Echocardiogram Classification
IEEE International Conference on Big Data and Smart Computing (BigComp), pp. 287–294
Interests
Research interests: Uncertainty quantification under distribution shift, Time-series representation learning under change points, Explainable AI for clinical decision support