LCD 검사 공정에서 가상 계측을 위한 머신 러닝 기반 예측 모델

송세리,박상철 한국CDE학회 2019 한국CDE학회 논문집 Vol.24 No.3

This paper proposes a machine learning based prediction model construction methodology for the virtual metrology in LCD manufacturing processes. The proposed prediction model construction methodology consists of four major steps; 1) data preprocessing, 2) feature selection, 3) deep learning model design, and 4) model validation. To extraction effective predictor variables at the feature selection stage, this paper employs three techniques including relative weight method, random forest method, and genetic algorithm. The constructed prediction model has been applied to LCD manufacturing data, and shows reasonably acceptable prediction accuracy which is higher than 90%.

산림재적 추정을 위한 계층적 베이지안 분석

송세리,박주원,김용구,Song, Se Ri,Park, Joowon,Kim, Yongku 한국데이터정보과학회 2017 한국데이터정보과학회지 Vol.28 No.1

산림경영 계획을 위한 필요한 산림재적을 보다 효율적으로 추정하기 위해서 다양한 연구가 요구되어져 왔는데, 이러한 산림구조에 관한 연구는 주로 현장조사와 위성영상을 이용하여 이루어진다. 현장조사를 통한 연구는 비교적 정확하나 시간과 비용이 많이 들 뿐 아니라 접근의 용이성이 떨어지는 지역이 있기 때문에, 넓은 지역의 조사가 어렵다는 단점이 있다. 최근에는 항공기에서 발사된 레이저 펄스가 반사되어 돌아오는 시간을 측정하여 대상의 3차원 좌표를 얻는 LiDAR (Light Detection and Ranging) 기술을 활용하여 획득한 정밀한 수치형자료를 이용한 산림의 구조에 관한 연구가 이루어지고 있다. 일반적으로 산림재적을 추정하기 위해서 LiDAR자료를 이용한 수고자료와 산림 재적에 대한 회귀모형의 중요성이 점차 높아지는데, 국내의 경우 수목의 종류와 그 분포가 다르기 때문에 회귀모형만으로 재적을 추정하는 데 한계가 있다. 따라서 본 논문에서는 산림의 수고와 흉고직경을 측정하여 재적값을 추정하고 산림의 공간효과를 고려한 계층적 베이지안 분석을 통해 관측되지 않은 전체 산림재적에 대한 추정을 하고자 한다. It has gradually become important to estimate a forest stand volume utilizing LiDAR data. Recently, various statistical models including a linear regression model has been introduced to estimate a forest stand volume using LiDAR data. One of limitations of the current approaches is in that the accuracy of observed forest stand volume data, which is used as a response variable, is questionable unstable. To overcome this limitation, we consider a spatial structure for a forest stand volume. In this research, we propose a hierarchical model for applying a spatial structure to a forest stand volume. The proposed model is applied to the LiDAR data and the forest stand volume for Bonghwa, Gyeongsangbuk-do.

전투기용 레이다 기반 SAR 영상 자동표적분류 기능 구조 및 CNN 앙상블 모델을 이용한 표적분류 정확도 향상 방안 연구

임동주,송세리,박범,Lim, Dong Ju,Song, Se Ri,Park, Peom 한국시스템엔지니어링학회 2020 시스템엔지니어링학술지 Vol.16 No.1

The fighter pilot uses radar mounted on the fighter to obtain high-resolution SAR (Synthetic Aperture Radar) images for a specific area of distance, and then the pilot visually classifies targets within the image. However, the target configuration captured in the SAR image is relatively small in size, and distortion of that type occurs depending on the depression angle, making it difficult for pilot to classify the type of target. Also, being present with various types of clutters, there should be errors in target classification and pilots should be even worse if tasks such as navigation and situational awareness are carried out simultaneously. In this paper, the concept of operation and functional structure of radar system for fighter jets were presented to transfer the SAR image target classification task of fighter pilots to radar system, and the method of target classification with high accuracy was studied using the CNN ensemble model to archive higher classification accuracy than single CNN model.

The effect of magnesium sulfate on postoperative pain in patients undergoing major abdominal surgery under remifentanil-based anesthesia

이철,장미순,송윤강,오세리,문서영,강동백,김병륜,변승재 대한마취통증의학회 2008 Korean Journal of Anesthesiology Vol.53 No.3

Background: Opioid tolerance may involve activation of the N-methyl-D-aspartate (NMDA) system. The possible involvement of the NMDA system suggests that one of the NMDA receptor antagonists, magnesium may be a useful adjunct to opioids for the treatment of postoperative pain following remifentanil infusion. Methods: For this study, 70 patients scheduled for major abdominal surgery under remifentanil-based anesthesia were randomly allocated into groups that received either magnesium sulfate (group M) or saline (group C) intravenously. The patients in the group M received 25% magnesium sulfate at a dose of 50 mg/kg in 100 ml of saline, and those in the group C received an equal volume of saline prior to the induction of anesthesia. In addition, patients in both groups received 10 mg/kg/h infusion of either magnesium sulfate (group M) or an equal volume of saline (group C) until the end of surgery. Pain was assessed using a visual analog scale at 30 min, and 6, 12, 24, and 36 hours after operation. The time to the first use of postoperative analgesic and cumulative analgesic consumption in both groups were also evaluated. Results: The visual analog scale scores for pain and cumulative analgesic consumption were significantly lower in the group M than in the group C. The time to the first use of postoperative analgesic was significantly shorter in group C than in the group M. Conclusions: Use of the NMDA-receptor antagonist, magnesium sulfate as an adjuvant analgesic reduced postoperative pain in patients undergoing major abdominal surgery under remifentanil-based anesthesia. Background: Opioid tolerance may involve activation of the N-methyl-D-aspartate (NMDA) system. The possible involvement of the NMDA system suggests that one of the NMDA receptor antagonists, magnesium may be a useful adjunct to opioids for the treatment of postoperative pain following remifentanil infusion. Methods: For this study, 70 patients scheduled for major abdominal surgery under remifentanil-based anesthesia were randomly allocated into groups that received either magnesium sulfate (group M) or saline (group C) intravenously. The patients in the group M received 25% magnesium sulfate at a dose of 50 mg/kg in 100 ml of saline, and those in the group C received an equal volume of saline prior to the induction of anesthesia. In addition, patients in both groups received 10 mg/kg/h infusion of either magnesium sulfate (group M) or an equal volume of saline (group C) until the end of surgery. Pain was assessed using a visual analog scale at 30 min, and 6, 12, 24, and 36 hours after operation. The time to the first use of postoperative analgesic and cumulative analgesic consumption in both groups were also evaluated. Results: The visual analog scale scores for pain and cumulative analgesic consumption were significantly lower in the group M than in the group C. The time to the first use of postoperative analgesic was significantly shorter in group C than in the group M. Conclusions: Use of the NMDA-receptor antagonist, magnesium sulfate as an adjuvant analgesic reduced postoperative pain in patients undergoing major abdominal surgery under remifentanil-based anesthesia.

Effect of early oral intake on postoperative bowel function in patients undergoing lower extremities surgery under epidural anesthesia

이철,이화성,송윤강,오세리,변승재,강동백,황지효,김병륜,문서영,한동엽 대한마취통증의학회 2008 Korean Journal of Anesthesiology Vol.53 No.3

Background: Early oral intake (EOI) associated with early recovery of normal bowel function has been shown to be an important determinant for improving patients' satisfaction. We investigated the tolerability of EOI and its effects on the recovery of bowel function after epidural anesthesia. Methods: A prospective randomized trial of patients undergoing lower extremities surgery under epidural anesthesia was performed. A liquid drink was given to 150 patients in the EOI group 1 hours after surgery, and to 150 patients in the delayed oral intake (DOI) group 8 hours after surgery. We recorded presence of bowel sounds immediately after operation, symptoms of ileus, time to the first flatus, time to the first defecation, degree of appetite before the first meal, and patients' satisfaction. Results: There was no significant difference in the presence of immediate postoperative bowel sounds, the degree of appetite before the first meal, mild ileus, and severe ileus between groups. Time to the first flatus and time to the first defecation in the EOI group were shorter than those of the DOI group. The patients' satisfaction in the EOI group was higher than that of the DOI group. Conclusions: For uncomplicated patients undergoing lower extremities under epidural anesthesia, beginning oral hydration as early as 1 hour after the operation is safe and well tolerated and resulting in faster recovery of bowel function and higher patients' satisfaction. Background: Early oral intake (EOI) associated with early recovery of normal bowel function has been shown to be an important determinant for improving patients' satisfaction. We investigated the tolerability of EOI and its effects on the recovery of bowel function after epidural anesthesia. Methods: A prospective randomized trial of patients undergoing lower extremities surgery under epidural anesthesia was performed. A liquid drink was given to 150 patients in the EOI group 1 hours after surgery, and to 150 patients in the delayed oral intake (DOI) group 8 hours after surgery. We recorded presence of bowel sounds immediately after operation, symptoms of ileus, time to the first flatus, time to the first defecation, degree of appetite before the first meal, and patients' satisfaction. Results: There was no significant difference in the presence of immediate postoperative bowel sounds, the degree of appetite before the first meal, mild ileus, and severe ileus between groups. Time to the first flatus and time to the first defecation in the EOI group were shorter than those of the DOI group. The patients' satisfaction in the EOI group was higher than that of the DOI group. Conclusions: For uncomplicated patients undergoing lower extremities under epidural anesthesia, beginning oral hydration as early as 1 hour after the operation is safe and well tolerated and resulting in faster recovery of bowel function and higher patients' satisfaction.

특허 요약의 밀도기반 클러스터링을 통한 미래 유망 기술 동향 파악 프로세스 (스마트 팩토리 분야 중심으로)

오유원(Yoowon Oh),송세리(Seri Song),박상철(Sangchul Park) 한국경영과학회 2021 한국경영과학회 학술대회논문집 Vol.2021 No.6

본 연구는 특허 내 요약부분을 밀도 기반 클러스터링 기법을 사용하여 여러 범주의 유망한 기술동향을 식별하는 방법론을 제안한다. 특허의 요약 부분에는 특허를 받고자 하는 내용과 발명의 특징 및 장점을 키워드로서 정리했기 때문에 이 안에는 해당 특허에 대한 핵심 기술 및 기술 트렌드의 동향이 함축되어 있다. 따라서, 우리는 많은 정보가 집약된 특허의 요약 부분을 활용하여 미래 유망 기술의 동향을 파악하고자 한다. 본 연구에서는, 현재 활발하게 연구가 이루어지고 있는 스마트 팩토리 분야의 기술 동향을 파악하고자 TF-IDF 기법으로 벡터화된 여러 키워드 중 관심 있는 분야와의 유사도 평가(코사인 유사도)를 진행하여 관심 기술에 대한 특허의 범위를 좁혔다. 이후, 밀도 기반 클러스터링(DBSCAN) 기법을 활용하여 카테고리 별 기술 동향을 파악하고 핵심 키워드를 뽑아내는 연구를 진행하였다.

특허 요약의 밀도기반 클러스터링을 통한 미래 유망 기술 동향 파악 프로세스 (스마트 팩토리 분야 중심으로)

오유원(Yoowon Oh),송세리(Seri Song),박상철(Sangchul Park) 대한산업공학회 2021 대한산업공학회 춘계학술대회논문집 Vol.2021 No.6

본 연구는 특허 내 요약부분을 밀도 기반 클러스터링 기법을 사용하여 여러 범주의 유망한 기술동향을 식별하는 방법론을 제안한다. 특허의 요약 부분에는 특허를 받고자 하는 내용과 발명의 특징 및 장점을 키워드로서 정리했기 때문에 이 안에는 해당 특허에 대한 핵심 기술 및 기술 트렌드의 동향이 함축되어 있다. 따라서, 우리는 많은 정보가 집약된 특허의 요약 부분을 활용하여 미래 유망 기술의 동향을 파악하고자 한다. 본 연구에서는, 현재 활발하게 연구가 이루어지고 있는 스마트 팩토리 분야의 기술 동향을 파악하고자 TF-IDF 기법으로 벡터화된 여러 키워드 중 관심 있는 분야와의 유사도 평가(코사인 유사도)를 진행하여 관심 기술에 대한 특허의 범위를 좁혔다. 이후, 밀도 기반 클러스터링(DBSCAN) 기법을 활용하여 카테고리 별 기술 동향을 파악하고 핵심 키워드를 뽑아내는 연구를 진행하였다.

증례보고 : Charcot-Marie-Tooth병 환자에서 발생한 Rocuronium에 의한 근이완에서의 회복 지연

정용관 ( Yong Kwan Cheong ),이철 ( Cheol Lee ),손용 ( Yong Son ),송윤강 ( Yoon Kang Song ),김태요 ( Tai Yo Kim ),김대중 ( Dea Jung Kim ),오세리 ( Se Ri O ) 대한마취과학회 2007 Korean Journal of Anesthesiology Vol.53 No.1

Charcot-Marie-Tooth disease, which is also known as hereditary motor and sensory neuropathy, is a heterogenous group of inherited diseases of the peripheral nerve. The spectrum of severity varies from asymptomatic individuals to those with severe limb abnormalities requiring corrective surgery. We report two brothers who had previously been diagnosed with Charcot-Marie-Tooth disease 3 years earlier and were scheduled to undergo a correction osteotomy of both feet under general anesthesia. General anesthesia was induced with propofol 2 mg/kg, rocuronium 0.8 mg/kg and was maintained with O2-N2O-Sevoflurane. The younger brother showed no delay in recovery of the neuromuscular blockade but the elder brother showed a delay. (Korean J Anesthesiol 2007; 53: 145~9)