거주자의 자연 환기 행위 예측을 위한 주성분분석

Abstract

The purpose of this study is to predict the window opening behavior to reduce the energy consumption in buildings. Window Opening & closing is most preferred behavior to adjust their indoor environment. However these behavior causes an increase in energy consumption in the building In the previous study, the environmental factors (temperature, humidity, etc.) were attempted to directly predict the window opening using machine learning techniques. Too many factors were applied to machine learning to try to make predictions, so I found that there was a problem in actually making prediction tools. Therefore, this study analyzed the flow of factors and principal component analysis to reduce the factors. Based on the result, the process of deleting factors or merging collinear factors will leave only the minimum number of factors to use in the prediction tool.