Residential Asset Pricing Prediction using Machine Learning

Residential asset price prediction and analysis are prevalent research topics in economy. Most researches focus on macroeconomy perspectives to explain the factors affecting residential asset prices. In this paper we examine some micro factors, like lot area, pool area, that can be used as features to predict house price. We fit a rather simple regression model which contains a few characteristics of a residential asset, and we are able to reach a fairly good result. Some machine learning algorithms such as random forest and support vector machine are also implemented to predict asset pricing. All regression models have a R squared over 0.9.