http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
- 中文 을 입력하시려면 zhongwen을 입력하시고 space를누르시면됩니다.
- 北京 을 입력하시려면 beijing을 입력하시고 space를 누르시면 됩니다.
RISS 인기검색어
- 검색키워드
- 저자 : 안제성 ( Jesung Ahn ) (검색결과 5 건)
- 무료
- 기관 내 무료
- 유료
-
-
클러스트내 분산을 최소화시키는 2단계 자율 분류 신경회로망
김종완(Jongwan Kim),안제성(Jesung Ahn),황희융(Heeyeung Hwang) 한국정보과학회 1993 한국정보과학회 학술발표논문집 Vol.20 No.1
임의의 패턴들을 자율적으로 유사한 패턴군들로 분류해주는 비지도 학습 신경회로망 모델을 제안한다. 제안한 방법은 안정된 클러스터 발견을 위해서 2단계 학습(2-phase learning)을 수행한다. 1단계에서는 클러스터 영역 임계치를 동적으로 변화시켜서 클러스터내 분산이 최소가 되도록 학습한다. 2단계에서는 학습된 뉴런들중에서 학습이 많이 되고 안정된 뉴런들을 1차 출력 뉴런으로 선정하고 나머지 뉴런들을 기선정 출력 뉴런들과의 거리와 학습 안정도에 따라 출력 뉴런에 병합시키거나 새로운 출력 뉴런으로 생성한다. 제안한 학습 알고리즘을 인공적으로 만든 데이타와 실제 음성 데이타에 적용하여 실험한 결과 우수한 성능을 보이면서 안정된 학습을 수행함을 확인하였다.
-
감소된 입력 차원을 가지는 경쟁 학습 신경망을 사용한 패턴 인식
김종완(Kim Jongwan),안제성(Ahn Jesung),김종상(Kim Chong Sang),조성원(Cho Seongwon),황희융(Hwang Heeyeung) 한국정보과학회 1994 정보과학회논문지 Vol.21 No.10
클래스 판정 방식을 갖는 다중 신경망을 사용하는 새로운 경쟁 학습 알고리즘을 제안한다. 기존의 신경망들은 주어진 입력 차원의 모든 성분들을 사용하여 학습하고 인식을 수행한다. 하지만 패턴 인식 문제들에 적용시켜 보면 모든 차원 성분들이 인식 작업에 필요한 것은 아니고, 어떤 경우에는 특정 차원이 인식을 어렵게 만드는 경우도 발생한다. 본 논문에서는 입력 차원 성분의 일부를 제거하여 학습시킨다. 우리는 기존의 경쟁 학습 알고리즘들과 달리 학습 종료후 애매한 출력 뉴런은 인식 작업에서 제외하므로써 인식률을 향상시켰다. 다중 신경망을 통합하여 인식 클래스를 결정하기 위하여 3가지의 판정 방식을 개발하였다. 원격 탐사 데이타를 사용한 실험에서 제안한 방법은 기존의 방법들에 비하여 우수한 결과를 보였다. A new competitive learning algorithm using an ensemble of neural networks with a consensus scheme is proposed. Conventional neural networks utilize all the dimensions of the original input patterns for learning and classification. However, a particular attribute of the input patterns does not necessarily contribute to classification and may even cause misclassification in certain cases. In this paper, we use the reduced input dimension for training a neural network. Unlike the conventional competitive learning algorithms, we eliminate ambiguous output neurons which cannot be assigned to any class in order to improve classification accuracy. We have developed three consensus schemes to judge the class using multiple neural networks. The experimental results with remote sensing data indicate the superiority of the proposed method in comparison to the conventional methods.
-
라이프스타일 요인에 따른 세부집단별 특성분석 - 항공서비스에 대한 함의 -
강명수 ( Myung Soo Kang ),이정훈 ( Jeonghoon Lee ),안제성 ( Jesung Ahn ) 한국항공경영학회 2020 한국항공경영학회지 Vol.18 No.1
The purpose of this study is to suggest the necessity of developing and providing customized aviation service for each cluster by grasping the characteristics of segmented clusters by clustering by lifestyle type in order to meet the needs of consumers who have fast and diversified aviation services. For this, a questionnaire was prepared based on theoretical considerations, and data were collected. Factor analysis and reliability analysis were conducted to secure the validity of the measured variables. The lifestyle factor analysis was divided into seven factors: appearance-oriented, social-oriented, family-oriented, brand/showoriented, leisure-oriented, goal-oriented, and social-oriented, and the result of cluster analysis was divided into four groups. A one-way analysis was conducted to identify differences between groups that were subdivided into different life styles. The results showed that there were significant differences between the subdivided clusters. The lifestyle attribute of each cluster was selected by factors with high positive values in each cluster. Based on this, the characteristics of each cluster could be identified. Cluster 1 enjoys various cultural life in life, makes human relations important, and cluster 2 has a strong sense of community and tends to regard human relations as important. Cluster 3 likes to get along with people and thinks about spending time with family, and cluster 4 has a clear intention in life and a tendency to value what is seen by others. As a result, it was found that life style was useful for market segmentation by identifying the attributes that were considered important for each cluster that were subdivided into different life styles. In addition, it was also possible to identify areas of importance in customer management by segmented clusters when using aviation services, and it is believed that the fragmented breakdown of characteristics according to lifestyle will help establish marketing strategies to meet the growing needs of air service users and also help them find directions for customer satisfaction.
-
편의체험과 쾌락적 체험 구성요인이 카페 매력도에 미치는 영향 - 기내식음료서비스 체험에 대한 함의 -
김재환 ( Jaehwan Kim ),이정훈 ( Jeonghoon Lee ),강명수 ( Myung Soo Kang ),안제성 ( Jesung Ahn ) 한국항공경영학회 2020 한국항공경영학회지 Vol.18 No.6
This study was initiated to improve the problems with the experience of in-flight food and beverage service. To this end, based on a study targeting users of cafe stores, which are representative examples of food and beverage services, we intend to derive an in-flight food and beverage service experience strategy that applies the empirical results of convenience and pleasure experiences. Specifically, the two basic dimensions of customer experience were presented as a convenience experience and a hedonic experience. The lower level of the convenience experience includes utility convenience and decision-making convenience, and the lower level of the hedonic experience includes emotional experience and relational experience. Through empirical verification of the two-dimensional hierarchical structure of customer experience presented in this study, the influence of convenience experience and pleasure experience factors on the attractiveness of cafes was identified. Of the 200 surveys distributed for analysis, a total of 175 pieces of data were used for the final analysis, except for 15 parts that were not collected or were unreliable. Screening and refinement of measured items, reliability analysis and feasibility analysis were conducted. As a result of factors analysis, the convenience experience factors consisted of utility convenience and decision-making convenience, and the hedonic experience factors consisted of emotional experience and related experiences. Correlation analysis and covariance structural model analysis was used for hypothesis testing in this study. As a result of the analysis, it was confirmed that utility and decision-making factors had a positive effect on cafe attractiveness, and emotional factors influenced the opposite direction for pleasure experience, and relational factors were not statistically significant. The results of this study are meaningful in providing information that can be helpful in the aspect of operating in-flight food and beverage services by identifying the differences in the attractiveness of customer experience factors and suggesting directions for future marketing strategy establishment.
KCI연관 검색어 추천
이 검색어로 많이 본 자료
활용도 높은 자료
