Open AccessMATRIX RECOVERY METHODOLOGY BASED ON MOBILE DATA
Author(s) -
Ol'ga Lebedeva
Publication year - 2021
Publication title -
sbornik naučnyh trudov angarskogo gosudarstvennogo tehničeskogo universiteta
Language(s) - English
Resource type - Journals
ISSN - 2686-7788
DOI - 10.36629/2686-7788-2021-1-1-117-120
Subject(s) - computer science , matrix (chemical analysis) , data mining , sample (material) , algorithm , chemistry , materials science , chromatography , composite material
The study is aimed at comparing the methods of restoring the correspondence matrices and a more detailed study of the algorithm for restoring the correspondence matrix based on innovative mobile technologies. Previously, the correspondence matrices were reconstructed by interviewing vehicle drivers and applying gravity models, then matrix estimation methods are used to calculate traffic flows. Each of these approaches has a number of advantages and disadvantages, but there is no evidence to assess the quality of the data obtained. The results show that, in general, processing da-ta from the mobile network allows for a more reliable estimate without taking into account survey data or similar observations, and using this type of research can lead to a more consistent assessment of trips with a larger sample size