http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
유도전동기 자가 진단 및 상수 추정을 위한 고주파 전류 제어기 구현
권영수(Young-Su Kwon),석줄기(Jul-Ki Seok) 대한전기학회 2007 대한전기학회 학술대회 논문집 Vol.2007 No.10
High frequency voltage signal injection have been widely used but they have some problems like over current protection. High frequency current signal injection and feedback control are more stable than voltage signal injection. In this paper, high frequency current controller for self-sensing and parameter estimation of induction motors is presented.
권영수(Young Su Kwon) 현대문법학회 2000 현대문법연구 Vol.20 No.-
Prototypensemantik und Stereotypensemantik haben zu zwei wesentlichen Veränderungen gegenüber der traditionellen Semantik geführt. Daher wird die in der Sprachwissenschaft bis dahin übliche Trennung von sprachlichem und außersprachlichem Wissen aufgehoben. Die notwendigen Merkmale werden durch prototypische oder stereotypische Merkmale ersetzt und sind vom Status her nicht mehr gleichwertig, sondern mehr oder weniger typisch. Obwohl beide Ansätze aus unterschiedlichen theoretischen Umfeldern stammen, werden sie oft wie terminologische Varianten zur Bezeichnung desselben Phänomens in Deutschland verwendet. Ihnen ist der Versuch gemeinsam, eine Erklärung für die Subsumption von Gegenständen unter einen Begriff zu geben. Aber es gibt grundlegende Unterschiede zwischen diesen beiden Theorien: im einen Fall geht es um den sprachlichen Bezug auf die Wirklichkeit, im anderen um die mentalen Prozesse der Sprachverarbeitung. Für die sinnvolle Verwendung der Ideen wird in der vorliegenden Arbeit versucht, die Rolle von Prototyp und Stereotyp in der lexikalischen Semantik und Unterschiede zwischen Prototypensemantik und Stereotypensemantik zu beleuchten.
오진영(Jinyoung Oh),차정원(Jeong-Won Cha) 한국정보과학회 2021 정보과학회논문지 Vol.48 No.5
딥러닝은 자연어처리 분야에서 우수한 성능을 보이고 있다. 하지만 우수한 성능을 달성하려면 많은 학습 데이터와 긴 학습 시간이 필요하다. 우리는 관계 추출 문제에 대하여 뉴럴-심볼릭 방법을 이용하여 적은 학습 데이터 환경에서 딥러닝의 성능을 능가하는 방법을 제안한다. 규칙 결과와 딥러닝 결과와의 불일치도를 사용하는 구조를 설계하였다. 또한 수렴속도를 향상시키기 위해서 논리 규칙 필터링을 제안하고 규칙의 성능을 높이기 위해 문맥을 추가하였다. 제안 구조는 적은 학습 데이터에 대해서 우수한 성능을 보였으며, 빠른 성능 수렴이 이루어지는 것을 확인하였다. Deep learning has been continually demonstrating excellent performance in the field of natural language processing. However, enormous training data and long training time are required to achieve good performance. Herein, we propose a method that exceeds deep learning performance in a small learning data environment by using a neural-symbolic method for the relationship extraction problem. We have designed a structure that uses the inconsistency between the rule results and deep learning results. In addition, logical rule filtering has been proposed to improve the convergence speed and a context has been added to improve the performance of the rule. The proposed method showed excellent performance for a small amount of training data, and we confirmed that fast performance convergence was achieved.
권영수(Kwon, Young Su) 언어과학회 2020 언어과학연구 Vol.0 No.95
According to Social Role Theory, gender stereotypes are characterized as dynamic constructs influenced by actual and perceived social role changes of men and women. When gender inequality in the economic sphere was persistent in Korea, a lower level of women’s participation was reflected in social roles. However, as the employment rate of women has steadily increased since the mid 80s, this change of social trend contributed to the dynamic aspects of gender stereotype. This study investigated how different age groups perceive gender stereotypes. It showed that the factor of age plays a significant role in the perception of dynamic gender stereotypes. It was found that the younger age group, consisting of teenagers, showed a significant discrepancy between the actual social employment conditions and the perception level of dynamic gender stereotypes. On the other hand, the psychodynamic perception of the older age groups appeared more clearly than in the other age groups.