In this work, the real estate market in Almaty city districts was considered. We know that this market is an important sector of the world economy. The residential real estate market in Kazakhstan is a complex structure consisting of hundreds of thousands of apartments characterized by many features. At the same time, any changes in the market can cause speculation and deliberate increase in real estate prices. That is why it is so important to understand what the real cost of an apartment is and where it is more expensive. In the course of the study, the creation of a real estate value prediction model was considered using machine learning methods. Several algorithms of machine learning: Linear Regression, Lasso, Ridge, Decision Tree Regression, Random Forest Regression, SVR (Gaussian kernel) were used in the work, their work results and accuracy were shown.