Analysis of Indicators of the State of Regional Freight Traffic by Method of Fuzzy Linear Regression

摘要

In this paper on the basis of statistical data 1996-2017 an analysis was made of the dependence of the volume of goods transported by road in the Rostov region on the density of public roads with hard surface, gross regional product per capita and tariff indices for freight traffic. When constructing regression equations on the basis of economic data, there is often an uncertainty that is associated with incomplete and vague information about the process being studied. In this reason, a fuzzy linear regression method is proposed for the analysis of this dependence. The linear regression coefficients are fuzzy symmetric triangular numbers. To find them, the corresponding optimization problem is solved. Two models were constructed, corresponding to the degrees of the fitting of the fuzzy linear model h = 0.4 and h = 0.5. The most adequate of the constructed fuzzy models is the model corresponding to the degree of the fitting of the fuzzy linear model h = 0.4, which is confirmed by the analysis of the control sample. At h = 0.5, the fuzziness coefficient increases sharply and the application of the model has no practical value. According to the results of the analysis, it was found that the index of tariffs for freight transportation has a decisive influence on the volume of cargo transportation.