머신러닝 분석을 통한 분양가 상한제의 주택시장 영향 연구

초록

Housing prices in Korea continued to rise with the consumer’s desire for asset increase and suppliers’ desire for profit, and the governmentintroduced a price ceiling system in 1972 to stabilize housing prices. The government expected that housing stability would be achieved byintroducing the price ceiling system, but there was a constant controversy over the effect of it. Some experts argued that the price ceilingsystem does not affect existing housing prices, but rather reduces housing supply, which increases price in the long run. In this study, weintend to analyze the impact of the price ceiling system on housing market forecast by predicting housing market based on the presence orabsence of the price ceiling system and comparing the accuracy. Several previous research have studied the factors affecting housing price,but this only identified the correlation, and there was insufficient of research on housing price prediction according to actual policy. Toovercome such statistical limitations, this paper predicts housing price using machine learning that can be interpreted with high accuracy evenif input variables are incomplete, wide, or irregular. As a result of analysis, the difference between RMSE(Root Mean Error), error ratemean, and error rate standard deviation was insignificant even if the price ceiling system was applied or not. This means that the housingmarket is more affected by other factors than the price ceiling system.

키워드