Application of a Nonparametric Pattern Recognition Algorithm to the Problem of Testing the Hypothesis of the Independence of Variables of Multi-Valued Functions

The problem of hypothesis testing for the independence of two-dimensional random variables in the analysis of variables of multi-valued functions is considered. To solve it, we used a technique based on a nonparametric kernel-type pattern recognition algorithm corresponding to the maximum likelihood...

Full description

Saved in:
Bibliographic Details
Published in:Measurement techniques 2022-04, Vol.65 (1), p.17-23
Main Authors: Lapko, A. V., Lapko, V. A., Bakhtina, A. V.
Format: Article
Language:English
Subjects:
Algorithms
Analytical Chemistry
Characterization and Evaluation of Materials
Hypotheses
Measurement Science and Instrumentation
Nonparametric statistics
Object recognition
Pattern recognition
Physical Chemistry
Physics
Physics and Astronomy
Random noise
Random variables
Remote sensing
Two dimensional analysis
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The problem of hypothesis testing for the independence of two-dimensional random variables in the analysis of variables of multi-valued functions is considered. To solve it, we used a technique based on a nonparametric kernel-type pattern recognition algorithm corresponding to the maximum likelihood criterion. The technique made it possible to bypass the problem of decomposing the random variable domain of values into intervals. Based on the results of computational experiments, the effectiveness of the applied technique was estimated depending on the type of multi-valued functions, the level of random noise and the amount of initial statistical data. The results obtained are relevant for solving the problem of detecting natural and technical objects from remote sensing data.
ISSN:0543-1972
1573-8906
DOI:10.1007/s11018-022-02043-2