개인구독
소속 기관이 없으신 경우, 개인 정기구독을 하시면 저렴하게
논문을 무제한 열람 이용할 수 있어요.
지원사업
학술연구/단체지원/교육 등 연구자 활동을 지속하도록 DBpia가 지원하고 있어요.
커뮤니티
연구자들이 자신의 연구와 전문성을 널리 알리고, 새로운 협력의 기회를 만들 수 있는 네트워킹 공간이에요.
논문 기본 정보
- 자료유형
- 학위논문
- 저자정보
- 지도교수
- 이경섭
- 발행연도
- 2020
- 저작권
- 영남대학교 논문은 저작권에 의해 보호받습니다.
이용수6
이 논문의 연구 히스토리 (2)
- 이 논문의 후속연구가 궁금하신가요?
- 연관 학술논문 또는 학술발표를 통해 보다 발전된 연구결과를 확인하실 수 있습니다.
- 이 논문의 연구 히스토리 확인하기
초록· 키워드
오류제보하기
In this paper, we developed an Auto-Valuation Model using Seoul apartment transaction
data and the latest machine learning models. It is well known that traditional HPM models
suffer from multicollinearity when developing automatic evaluation models. We developed
an automatic evaluation model using random forest, XGBoost, LightGBM, and Stacking
models, which are nonlinear models and the latest machine learning models. We trained and
tested model by using the data that a total of 710,731 Seoul apartment transaction data from
the Public Data Portal from January 2010 to December 2018. Comparing the quartiles and
MAPE showed the superiority of the tree-based machine learning models, and important
variables were USM, Latitude, and Longitude. These analyzes confirmed the importance of
variables related to locations that could not be modeled in the existing HPM model, and
we can conclude that tree-based machine learning models, which can effectively avoid the
multicollinearity while modeling these variables, relatively explain data better.
data and the latest machine learning models. It is well known that traditional HPM models
suffer from multicollinearity when developing automatic evaluation models. We developed
an automatic evaluation model using random forest, XGBoost, LightGBM, and Stacking
models, which are nonlinear models and the latest machine learning models. We trained and
tested model by using the data that a total of 710,731 Seoul apartment transaction data from
the Public Data Portal from January 2010 to December 2018. Comparing the quartiles and
MAPE showed the superiority of the tree-based machine learning models, and important
variables were USM, Latitude, and Longitude. These analyzes confirmed the importance of
variables related to locations that could not be modeled in the existing HPM model, and
we can conclude that tree-based machine learning models, which can effectively avoid the
multicollinearity while modeling these variables, relatively explain data better.
목차
1. 서론 12. 데이터 소개 42.1. 데이터 수집 42.2. 데이터 탐색 52.2.1. 상관관계 62.2.2. 위치 73. 연구방법 83.1. 모형 소개 83.1.1. Random Forest 모델 93.1.2. XGBoost 모델 103.1.3. LightGBM 모델 113.1.4. Stacking 모델 133.1.5. Comparable Market Analysis 153.2. 예측분석 154. 실증분석 174.1. 분석 결과 174.2. 변수 중요도 235. 토의 및 결론 28참고문헌 29
최근 본 자료
댓글(0)
0