Estimation of Commercial Property Premiums Using Machine Learning : Focusing on Comparative Analysis Between Models

Although commercial property premiums in Korea are appraised based on appraisal practice standards, subjective factors cannot be completely excluded; this leads to disputes between landlords and tenants. Thus, empirical research on estimating commercial premiums using objective and quantitative factors should be conducted so that subjective factors can be excluded from the premium appraisal. Nevertheless, finding empirical research about estimating commercial premiums using only objective and quantitative formative factors is challenging . Herein, we conducted an empirical study to develop a model that can estimate commercial premiums using only objective formative factors using machine learning and validated the statistical significance of the suggested machine learning model by comparing the multivariate linear regression model. Results contribute to the academic area by empirically proving the effectiveness of the machine learning model and commercial premium market by introducing transparency and fairness to premium-minimizing disputes.