Year: 2021
Type: conference-paper
Abstract: The paper describe a solution of the problem of increasing the accuracy of calculating potential transport and pedestrian activity in the context of a shortage of marked data by using a geosocial approach. The advantages of the proposed model are the presence of a high number of degrees of freedom, which allows to flexibly configure the model based on the problem being solved. The authors note that the application of the approach according to the methodology proposed in the article to solve the problem of improving the efficiency of classifying pedestrian transport network data requires an individual project approach to the formation of initial data for analysis. We formulate a hypothesis that the accuracy of the forecast of transport demand and supply can be increased if the classifying model takes into account and analyses not only the properties of a particular territory, but also the characteristic features of the systems with which it interacts. Hexagons and heat maps were chosen for the classification of territories. A feature of the proposed approach is the unification and simplification of the neural network architecture..
Authors: Yu. A. Lezhnina, N. S. Maltseva
Language: English
Subfield: Transportation
Field: Social Sciences
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W4200470599