원문정보
A Study on the Validity of Dimensional Reduction Modelling for Making Architectural Space Information Tangible
초록
영어
The purpose of this study is to verify of an objective process for making architectural space information tangible. Research method is as follows. First, Present modelling techniques that reduce the dimensions of architectural spatial information. Second, Extract data from the observation group with dimensional reduction modelling techniques and then apply it to the cluster analysis method to demonstrate its validity. Third, Apply to the comparison data to verify its validity by comparing and aggregating it with the observation group. The results of the study are as follows. Dimensional reduction modelling techniques were proposed to enable cluster analysis, consisting of lightweight configuration and attribute information based on the physical elements of the building. Cluster analysis was applied so that the building could be clustered according to the characteristics of the building. In addition, the consistency of dimensional reduction modelling techniques was verified in comparison to the probability density function. In the future, studies should continue to be conducted on the types of large paper drawings provided by the National Archives of Korea.
목차
1. 서론
1.1 연구 배경 및 목적
1.2 연구 방법과 대상
1.3 차원 축소 모델링의 용어 정의
2. 선행 연구와 분석 도구 고찰
2.1 선행 연구 고찰
2.2 추출한 지수의 분석 도구 검토
3. 차원 축소 모델링 기법의 정의
3.1 차원 축소 모델링의 정의
3.2 차원 축소 모델링의 적용
3.3 클러스터 분석에 의한 군집
3.4 차원 축소 마이닝의 연속성 검증
4. 차원 축소 모델링의 신뢰성 논증
4.1 비교군의 설정과 적용
4.2 비교군 적용 결과의 타당성 검증
4.3 관찰군과 비교군의 종합 분석
5. 결론
REFERENCES