Youngko, Hyung
Pohang University of Science and Technology
研究室紹介
Professor Youngko's research lab specializes in data-driven systems and intelligent modeling, focusing on the integration of big data, machine learning, and statistical analysis to solve real-world challenges in healthcare, energy, aquaculture, and infrastructure. The lab develops advanced predictive models—particularly using probabilistic and state-space methods—for optimizing system performance and decision-making under uncertainty. Key research directions include e-health strategy development, energy consumption analysis via smart metering, and the application of Gaussian process regression for sustainable aquaculture. The lab also explores innovative AI techniques such as Retrieval-Augmented Generation with controlled noise to enhance language model performance.
Research Overview
Research Output Trend
Figures are computed from collected data and may differ slightly.
Selected Papers
6The adequacy of the urologist work force in Korea has never been investigated. This study investigated the geographic distribution of urologists in Korea. County level data from the National Health Insurance Service and National Statistical Office was analyzed in this ecological study. Urologist density was defined by the number of urologists per 100,000 individuals. National patterns of urologist density were mapped graphically at the county level using GIS software. To control the time sequenc
A well-established e-health strategy at the national level is necessary to successfully achieve the trust-and-consensus- based e-health goals by linking strategic information planning and the execution of an implementation plan. This paper provides a methodology of how to establish a national e-health strategy and the case of e-Health Information Strategic Planning (ISP) of the Ministry of Health and Welfare of Korea. The ISP is to improve the quality of care and contribute to the economic growt
최근 수산 자원의 고갈에 따른 육상 양식장에서의 ‘기르는 어업‘에 의한 생산성 향상에 대한 기대가 크게 고조되고 있다. 육상 양식장의 경우, 해상 환경과 달리 환경 및 양성 요소에 대한 제어와 관리가 용이하며, 출하 계획에 따른 생산량 조정이 가능한 이점이 있다. 반면, 자연 환경에서와 달리 어류 성장을 위한 인위적인 관리가 필요하기 때문에 운영에 따른 비용이 크게 증가할 수 있는 단점이 있다. 따라서, 계획된 목표 출하량에 맞추어 효율적으로 양식장을 운영함으로써 이윤 극대화를 추구할 수 있다. 이러한 효율적인 양식장 운영 및 어류 양성을 위해서는 대상 어종에 따른 정확한 성장 예측 모델이 반드시 요구된다. 현재까지 대부분의 성장 예측 모델은 양식장 수집 데이터를 활용하여 통계적 분석 기반의 수치 해석적인 결과들이 주를 이룬다. 본 논문에서는 기존의 통계적 관점에 의한 성장 예측 모델이 가질 수 있는 데이터 확보의 어려움 및 낮은 정확도에 대한 정량적 수치를 제공하기 어려운 단점을
Purpose: This paper introduces Pohang University of Science Technology (POSTECH) advanced metering infrastructure (AMI) and Open Innovation Big Data Center (OIBC) platform and analysis results of electricity consumption data collected via the AMI in POSTECH campus. Methods: We installed 248 sensors in seven buildings at POSTECH for the AMI and collected electricity consumption data from the buildings. To identify the amounts and trends of electricity consumption of the seven buildings, electrici
Purpose: This study investigates the state-space explosion problem in discrete-time Markov chain (DTMC) models of circular k-out-of-n: G balanced systems under cumulative shock-induced system failures and develops a systematic state-space compression procedure based on transition-structure equivalence. Methods: States are layered according to the number of failed units, and sub-DTMCs within the same layer that have identical reachable structures and transition probabilities are merged. Transitio
Retrieval-Augmented Generation (RAG) has emerged as a technique for improving the performance of language models (LLMs), where the inclusion of noisy documents is unavoidable. While prior studies have largely treated noise as detrimental and focused on its removal, this study demonstrates that certain types and combinations of noisy documents can improve LLM-based question answering performance. We categorize noisy documents into six types and systematically evaluate their individual and combine