https://scholar.dgist.ac.kr/handle/20.500.11750/113 2026-04-05T13:58:05Z https://scholar.dgist.ac.kr/handle/20.500.11750/59189 Title: DEVICE AND METHOD FOR CALCULATING STROKE VOLUME USING AI Author(s): 이형철; 양현림; 정철우; 김민수 Abstract: The present invention contains a stroke volume calculation device using AI, comprising: a filtering unit for filtering a stroke volume and an artery blood pressure value that are in a preset range, from first data and second data comprising a plurality of artery blood pressure values and a stroke volume corresponding to each artery blood pressure value; a pre-training unit for pre-training, by using third data filtered from the first data, a first stroke volume calculation model for calculating a stroke volume on the basis of the artery blood pressure value; a transfer learning unit for generating a second stroke volume calculation model by transfer learning, by using fourth data filtered from the second data, the first stroke volume calculation model for calculating a stroke volume on the basis of the artery blood pressure value; and a stroke volume calculation unit for calculating a stroke volume corresponding to inputted artery blood pressure of a specific patient by using the second stroke volume calculation model. https://scholar.dgist.ac.kr/handle/20.500.11750/57174 Title: 그래프 처리 시스템 및 그래프 처리 시스템의 동작 방법 Author(s): 김민수; 안규현; 박힘찬; 오세연; 김진욱 Abstract: 일 실시예에 따른 그래프 처리 시스템은 위상 데이터 및 속성 데이터를 포함하는 그래프 데이터를 저장하는 적어도 하나의 보조 기억 장치, 그래프 데이터의 일부를 저장하는 메인 메모리, 메인 메모리로부터 수신한 그래프 데이터의 처리 및 동기화를 수행하는 코어들 및 장치 메모리들을 포함하는 복수의 그래픽 처리 장치들, 및 복수의 그래픽 처리 장치들에서 수행되는 그래프 데이터에 대한 질의 처리를 관리하고, 질의 처리 결과 중 갱신 가능한 속성 데이터를 적어도 하나의 보조 기억 장치에 저장하는 중앙 처리 장치를 포함한다. https://scholar.dgist.ac.kr/handle/20.500.11750/47777 Title: A BERT-enhanced Graph Neural Network for Knowledge Base Population Author(s): Lim, Heechul; Kim, Min-Soo Abstract: We present BGKBP, a deep-learning algorithm based on BERT, and a graph neural network for knowledge base population (KBP). Our experiments showed that a straightforward application of BERT and GNN on a large knowledge base (e.g., Wikidata) improves KBP quality and outperforms the previous state-of-the-art methods. We developed four techniques to improve the BGKBP's KBP capability: (1) serialization, (2) fine-tuning, (3) node regression, and (4) text augmentation. BGKBP achieved the best F1 scores of 0.723 and 0.495 on entity linking and new entity detection, respectively. We further showed that using text augmentation (BGKBP-TA) achieved the best F1 score of 0.547 on relation linking, which is more difficult than entity linking because of the various representations of some of the relations. © 2023 IEEE. 2023-02-13T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/47123 Title: Toward Mission Critical A.I. Systems Author(s): Kim, Min-Soo 2017-10-18T15:00:00Z