Land Value Prediction Model in the Urban Fringe Region Using Multiple Regression Algorithm

Abstract

The existence of land has become a dream. The value continues to soar over the years, making the land become something that various parties highly desire. This phenomenon causes the demand for land also to continue to increase. This study aims to predict land values based on actual land prices in the field in the form of transaction results and bid prices with a statistical approach for mass valuation in the urban fringe area, where this area is considered an alternative area to overcome the need for housing which is getting higher and higher. Land value is estimated using a data comparison approach market and added to the cost of building construction. Multiple regression is one of the methods to get the best estimation model from big data. The magnitude of the contribution of the influence of the variables making up the Land Value can be known through the coefficient of determination (R2), which is 82.5%. Many factors affect the Land Value, such as Bid Price, Physical Land Factors, Environmental Factors, and Legal Aspects.