The article proposes an ensemble of machine learning algorithms and program results, including such big data management methods as regression, classification and clustering. The proposed methods in comparison allow to analyze and interpret the obtained data with the real circumstances in the real estate market. Information about real estate in the capital of Kazakhstan is considered as data. Big data is structured by such fields as cost, class, kitchen space size, area and is presented as a .csv file, processed using machine learning methods. Python was used as a programming environment, while the numpy, pandas, matplotlib, Axes3D, LinearRegression, Scikit-learn, KMeans libraries allow you to interpret and visualize the received data. The conducted computational experiment clearly demonstrates the classification of data, division into clusters, and also forms a forecast for the cost depending on the declared features.