| Abstract [eng] |
The main aim of this master thesis is to analyze and evaluate the real estate market using machine learning. The work consists of three main parts: the analysis of literature, the description of empirical research process, its methods and results. In addition, the conclusion is presented. In the empirical research the comparison of the results of macroeconomics indicators such as Gross Domestic Product, Unemployment rate and other indicators were carried out in the period of 2008 – 2018. The prognosis of these indicators is calculated till 2022. The analysis of the literature provides valuation and a deep learning prediction of real estate market. The main aim of this paper is by developing the machine learning method and automatic value estimation to identify the varieties of the price in the real estate regarding the street and the region. The model is going to be useful for real estate investors who want to predict the price. The neural network model is presented by doing prediction of large amount of databases. The housing market is vulnerable to price fluctuations regarding available correlations with many factors and variables that have the most impact on the price of the unit. These variables should be taken into consideration. The final prediction of the research is successfully done with the help of software. The real estate price prediction using advanced technologies, such as statistical modelling using MySQL server for data storage and machine learning in real estate valuation is new and innovative for real estate investors. |
