COMPARISON OF VARIANCE AND REGRESSION APPROACHES ON EXAMPLE OF URBAN TRAFFIC DATA ANALYSIS
Abstract
Purpose. Analysis of variance and regression analysis are ones of the most important methods of mathematical and applied statistics. The purpose of the article is to present both approaches specifying their similarities and significant differences. Methods. Mathematical and statistical methods have been used to achieve the set purpose. Results and their discussion. The analysis of variance and regression analysis are demonstrated on the example of solving the problem of time of day-dependent simulation of average traffic speed over the Glazkov Bridge in the city of Irkutsk. For this purpose a single factor analysis of variance and a corresponding regression analysis with dummy variables have been carried out. Introduction of a new qualitative explanatory variable -weekday/day off allowed to conduct a two-factor analysis of variance and built a corresponding regression model. The use of the factor “general level of traffic congestions on Irkutsk roads” allowed to construct a regression model describing the effect of both qualitative and quantitative factors on the average travelling speed of traffic over the Glazkov Bridge. Conclusions. The similarity and difference of the variance and regression approaches to the analysis of data have been shown on the specific example. The results obtained through the analysis of variance can be reproduced in the analysis of the regression model where the influence of qualitative factors is described using dummy explanatory variables.
About the Authors
G. D. Gefan
Irkutsk State Transport University
Russian Federation
M. P. Bazilevsky
Irkutsk State Transport University
Russian Federation
References
1. Доугерти К. Введение в эконометрику; 3-е изд. М.: ИНФРА-М, 2009. 465 с.
2. Дрейпер Н., Смит Г. Прикладной регрессионный анализ; пер. с англ.; 3-е изд.; в 2 кн. М.: ИД «Вильямс», 2007. 912 с.
3. Себер Дж. Линейный регрессионный анализ; пер. с англ. М.: Мир, 1980. 456 с.
4. Как работают Яндекс. Пробки [Электронный ресурс]. URL: https://yandex.ru/company/technologies/yaprobki/ (25.11.2017).
5. Вучков И., Бояджиева Л., Солаков Е. Прикладной линейный регрессионный анализ; 2-е изд.; в 2 кн. М.: Финансы и статистика. Кн. 1 - 1986; кн. 2 - 1987. 239 с.
6. Мостеллер Ф., Тьюки Дж. Анализ данных и регрессия; пер. с англ.; в 2 вып. М.: Финансы и статистика, 1982. Вып. 1 - 349 с.; вып. 2 - 239 с.
7. Гефан Г.Д. Применение корреляционно-регрессионного анализа как направление исследовательской деятельности при обучении эконометрике // Современные технологии. Системный анализ. Моделирование. 2015. № 4 (48). С. 92-97.
For citations:
Gefan G.D.,
Bazilevsky M.P.
COMPARISON OF VARIANCE AND REGRESSION APPROACHES ON EXAMPLE OF URBAN TRAFFIC DATA ANALYSIS. Proceedings of Irkutsk State Technical University. 2018;22(1):58-68.
(In Russ.)
Views:
161