House Price Prediction in US Metropolitan Areas

For several decades, theorists and researchers have emphasized investigating and analyzing the likelihood of contagion of turbulence among markets. The behavior of the housing market is also essential because of the impact of housing prices on the loan portfolios of banks and other financial institutions. In the housing sector, it is expected that an expansionary financial and monetary policy will increase the demand for housing by increasing the amount of money in the asset portfolio. However, it depends on so many factors. Therefore, it is necessary to consider this matter profoundly. In recent years, the price prediction process in the housing market has also faced changes with the development of artificial intelligence, and various algorithms have been proposed for it. This article presents a combined method for selecting practical price prediction features. In the proposed model of the Harris Hawks optimization algorithm, the features affecting the price are selected by combining the random forest algorithm, and the forecasting process is carried out. Examining the results showed that the proposed model with 20 hawks provides its best performance. In such a way that of 14 features of housing, ten features are selected for the prediction process, and the accuracy is 63.70. In addition, compared to similar algorithms such as decision trees, nearest neighbor, and multilayer perception neural networks, the provided prediction error is less