Review article

The article considers the possibility of applyingensemble algorithms of decision trees in the problemsof mass evaluation of real estate, including cadastralevaluation. The results obtained by ensemble methods arecompared with the results of multiple linear regressionmodels. It is shown that machine learning methods basedon decision trees can be more accurate than regressionmodels. They can be used for verification of modelsbuilt by traditional evaluation techniques, as well as forproviding repeatability and reproducibility of results.