XGBOOST 모형을 활용한 부동산 조각투자 가격 예측: 카사 TE 물류센터를 중심으로

This study aims to investigate the price prediction of a real estate fractional investment platform using the deep learning XGBOOST model. In February 23, the Financial Services Commission announced the “Plan to Improve the Regulatory System for Issuing and Distributing Security Tokens”. In addition, on July 28, 2023, Yoon Chang-hyun, a member of the National Assembly, proposed a bill to amend the Electronic Securities Act and the Capital Market Act to apply to the capital market and financial investment business. Therefore, this study explains the definition of token securities and the current status of domestic policies and operations, and attempts to predict the price of real estate fractional investment in the “TE Logistics Center” product of Kasa, a representative real estate fractional investment platform. The empirical analysis using the XGBOOST model shows that the Train RMSE is 0.00424 and the Test RMSE is 0.03228. This means that the error rate is low and the predictive power is high, and we can see that the specific important variables that affect the price prediction are the closing price and the opening price. This confirms the effectiveness of deep learning models in predicting the price of real estate fractional investments, and we believe that by applying deep learning XGBOOST models to new investment markets such as real estate fractional investment platforms, we can provide investors with more reliable prediction information.