Title:

Using Rank Tests and Threshold Copulas for Classification ofMultidimensional Data Sets

Author:

Haroutunian v

Type:

Conference

Co-author(s) :

Safaryan Irina ; Harutyunyan Narine

Uncontrolled Keywords:

Change-point problem ; change-point detection techniques ; threshold dependence ; copula ; rank score test ; categoricalvariable

Abstract:

This report deals with classification of multidimensional data sets into statistically homogeneous groups and emphasizes the copula representation of the dependence between the components of a random vector. We present a nonparametric algorithm based on rank score test which allows reducing the investigation of changes in the joint distribution of a random vector components to investigation of some one-dimentional conditional distributions. Some applied examples are presented.

Language:

English

URL:


Additional Information:

eghishe@sci.am ; irinasafaryan@yandex.ru ; narineharutyunyan57@gmail.com

Affiliation:

Institute for Informatics and Automation Problems

Country:

Armenia

Year:

2015

Time period:

September 28 - October 2

Conference title:

10 th International Conference on Computer Science and Information Technologies CSIT 2015

Place:

Yerevan

Participation type:

oral