Abstract
This article examines existing scientific approaches to assessing and forecasting inflation based on various macroeconomic parameters. Differentiated inflation forecasting models and the main regressors used for such modeling are analyzed. The author concludes that it is necessary to refine approaches to forecasting inflation in Russia by using machine learning methods and adding a pool of qualitative variables to the model that more accurately characterize current market participants' expectations regarding the formation and dynamics of inflation in Russia based on the current geopolitical and macroeconomic situation. A future direction for the author's research is the development of a model for estimating the inflation risk premium used to determine the size of the premium.