The paper provides a brief analysis of publications on the application of regression analysis methods for railway transport. The problem of constructing a regression model of loading on the Krasnoyarsk railway, which is used as an autonomous model, is considered: the operated fleet of locomotives, unloading, the working fleet of railroad cars, the carrying capacity of the railway line. As an information base, we use the statistical information of the Krasnoyarsk railway for 2000-2018. The estimates of the parameters obtained using the method of mixed estimation (MES) are given. They are compared with the estimates calculated using the methods of least squares, moduli and anti-robust estimation traditional for regression analysis. It is concluded that the MES allows one to obtain a model that surpasses the others in terms of the consistency of behavior.

regression analysis, model, railway transport, method of mixed parameter estimation.