Русский
!

Conference publications

Abstracts

XXV conference

Research of a variational series according to the market of used cars

Perminova N.

Vyatka State University, institute of Mathematics and information systems, f-t. Computer and physical and mathematical sciences, dep. of Mathematical modeling, Russia, 610998, Kirov, Moskovskaya St., 36, E-mail: ns.perminova@yandex.ru

1 pp. (accepted)

In real socially - economic systems it is impossible to make active experiments therefore data usually represent observations of the happening process. Results of observations is a number sequence, located in a chaotic order which for studying needs to be arranged (to range).

Data for a research were taken from the Kaggle platform. An object of a research of work is the market of used cars. For the analysis the decision to use a number of distribution of a run of cars was made.

Research objective – check of a hypothesis that is the cornerstone of empirical distribution of size of car mileage the law of normal distribution. Check of the made hypothesis is carried out by means of criteria of consent. For calculation of characteristics of a variational series of distribution the program R – Studio is used.

In social and economic statistics normal distribution meets seldom, but comparison with it is important for clarification of degree and character of a deviation from it the actual distribution.

In this work of value of a median (144406.8) and mode (146375) are sufficient are close, in difference from average arithmetic size (123347.3), it indicates probable discrepancy of the studied distribution to the normal law.

On Pearson's formula χ2 =2458100, that there is more tabular value χ2(tabl)=35.72 at significance value α = 0,05 and number of degrees of freedom ν=17, means it is possible to say that the law of normal distribution isn't the cornerstone of empirical distribution of size of a run of cars.

As a result of a research, the made hypothesis of compliance of empirical distribution of size of a run of cars to normal distribution was disproved. Thereof, for the further analysis the hypothesis of distribution with "heavy tails" to which Pareto's distribution belongs can be made.



© 2004 Designed by Lyceum of Informational Technologies №1533