On segmentation approach for time series of arbitrary nature A. Sherstobitova, T. V. Emelyanova
Material type:
ArticleContent type: Текст Media type: электронный Subject(s): модели временных рядов | структурные разрывы | аномальное наблюдениеGenre/Form: статьи в сборниках Online resources: Click here to access online
In:
Applied methods of statistical analysis. Statistical computation and simulation : proceedings of the international workshop, 18-20 September 2019 P. 510-514Abstract: It is considered the problem of splitting the time series Xt of arbitrary nature
into segments generated by one generation mechanism, as well as detecting the
moment of change of one generation mechanism to another. Based on the
parameters of complexity, there is considered the time series segmentation
methodology, which does not require any a priori knowledge of the mechanisms
for their implementation.
Библиогр.: 2 назв.
It is considered the problem of splitting the time series Xt of arbitrary nature
into segments generated by one generation mechanism, as well as detecting the
moment of change of one generation mechanism to another. Based on the
parameters of complexity, there is considered the time series segmentation
methodology, which does not require any a priori knowledge of the mechanisms
for their implementation.
There are no comments on this title.