구성원
구성원
-
Hye-Young Jung 정혜영 교수
DepartmentProf. Dept. of Mathematics & Data Science, College of Science and Convergence Technology / Professor
E-mailhyjunglove@hanyang.ac.kr
Office Phone031-400-5463
Homepagehttp://mfmd.hanyang.ac.kr
- 학과/직책과학기술융합대학 수리데이터사이언스학과 / 교수
- 이메일hyjunglove@hanyang.ac.kr
- 전화번호031-400-5463
- 홈페이지http://mfmd.hanyang.ac.kr
- Research Keywords
- 연구키워드
- #Data Science #Machine Learning #Deep Learning #AI #Fuaay Modeling #Uncertainty quantification #Data Mining #Statistical modeling
- #데이터사이언스 #머신러닝 #딥러닝 #인공지능 #불확실성 정량화 #퍼지모델링 #퍼지이론 #소프트컴퓨팅
- Research Objectives
- 연구목표
- A study to build a reliable model through uncertainty quantification
- Developing measurements and models for explainable and reliable AI implementations
- Industry-related research that uses data to derive industry insights and predict the future
- 불확실성 정량화를 통해 신뢰성 있는 모델 구축을 위한 연구
- 설명가능 인공지능 및 신뢰가능한 인공지능 구현을 위한 측도 개발 및 모델 개발
- 데이터를 이용하여 산업의 인사이트를 도출하고 미래를 예측하는 산업관련 연구
- Brief Research Experience
- 주요경력
- Top 10% journals main author: Information sciences, fuzzy sets and systems
- National Research Foundation of Korea(2017.03 ~ 2019.02) Study on the use and usefulness of fuzzy set theory in precision medicine
- National Research Foundation of Korea(2019.06 ~ 2022.05) Study on the applications of uncertainty theories in medical diagnosis and bioinformatics
- National Research Foundation of Korea(2022.06 ~ 2025.02) A Study on Uncertainty Quantification in AI-based Decision Making
- 박사후연구원, 서울대학교 통계학과(2014-2016)
- 강의교수, 서울대학교 기초교육원 통계분야(2016-2019)
- 논문 SCI 25편 게재
- 한국연구재단 신진 연구, 창의도전 연구 수행 완료, 현재 기본연구 수행중(2022-2025)
기본연구 주제: 인공지능 기반 의사결정에서의 불확실성 정량화에 관한 연구
- Research Areas
- 연구분야
- Uncertainty quantification
- Explainable and reliable AI
- Modeling based on uncertainty theory
- 데이터가 지닌 불확실성 및 모델의 불확실성 정량화 연구
- 인공지능의 설명 가능성 및 신뢰성 연구
- 불확실성 이론에 기반한 인공지능 모델링 연구연구목표
- Thesis
- 논문
- Jung, H. Y., Leem, S., & Park, T. (2018). Fuzzy set-based generalized multifactor dimensionality reduction analysis of gene-gene interactions. BMC medical genomics, 11(2), 32.
- Jung, H. Y., Choi, H., & Park, T. (2018). Fuzzy heaping mechanism for heaped count data with imprecision. Soft Computing, 22(14), 4585-4594.
- Choi, S. H., Kim, H. S. & Jung, H. Y. (2019). Ridge fuzzy regression model. International Journal of Fuzzy Systems, 21(7), 2077-2090
- Lee, W. J., & Jung, H. Y. (2019). Statistical inference for time series with non-precise data. International Journal of Approximate Reasoning
- Yun, Y., & Jung, H. Y. (2021). Impacts of public medical insurance reforms on households: An application of fuzzy cognitive map for scenario evaluation. Soft Computing, 25(12), 7947-7956
- Min, H. J., Shim, J. W., Han, H. J., Park, C. H., & Jung, H. Y. (2022). Fuzzy Transform and Least-Squares Fuzzy Transform: Comparison and Application. International Journal of Fuzzy Systems, 24(6), 2740-2752
- Research Topics
- 연구내용
- 1. Uncertainty quantification
Research on techniques to define and quantify uncertainty that can occur in a model, research on measures and methods for measuring uncertainty in deep learning, and research on model development that reflects uncertainty by combining fuzzy and uncertainty theories
- 2. Explainable and reliable AI
Developing measurements and models for explainable and reliable AI implementations
- 3. Mathematical and statistical modeling for various industries
: Modeling for analysis and prediction of financial, bio and social and natural phenomena
Propose a unified estimator for the autoregressive model with fuzzy data based on the least squares method and investigate the optimality properties
Propose the estimation algorithm using the rank transform method which is known to be neither dependent on the shape of the error distribution nor sensitive to outliers
Solve the uncertainty of simple binary classification.
Propose a Fuzzy MDR which takes into account the uncertainty of binary classification by allowing the possibility of partial membership.
Detect significant gene-gene interaction through the real application
Propose the lasso fuzzy regression model, combining the concept of fuzzy set to represent imprecise data, with the lasso regression model
Employ fuzzy cognitive map to analyze how uncertainty shocks affect the household’s consumption,workinghours,andincomesources.
A Study on Uncertainty according to Outlier in deep learning model
A Stud on the similarities and differences between fuzzy transform and least-squares fuzzy transform, which are well known for re- constructing original function or removing noise.
Propose Least Absolute Deviation Fuzzy Transform(LAD-FT) algorithm, applied LAD(LeastAbsoluteDeviationApproximation)methodthatisrobusttooutliertoFuzzy transform.
A Study of SHAP and Feature selection in Statistical View
Noise robustness analysis of Shapley value for Deep SHAP
