ASSESSING THE PERFORMANCE QUALITY OF MULTICRITERIA NEURAL NETWORK STATISTICAL ANALYSIS USING THE EXAMPLE OF TESTING THE HYPOTHESIS OF INDEPENDENCE OF SMALL SAMPLE DATA

Background. The purpose of the article is to create a quality criterion for assessing the effectiveness of neural network processing of small samples in multicriteria testing of the independence hypothesis. Materials and methods. To date, about 200 statistical criteria for testing the hypothesis o...

Full description

Saved in:
Bibliographic Details
Main Authors: A.I. Ivanov, A.I. Ermakova
Format: Article
Language:English
Published: Penza State University Publishing House 2024-11-01
Series:Надежность и качество сложных систем
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Background. The purpose of the article is to create a quality criterion for assessing the effectiveness of neural network processing of small samples in multicriteria testing of the independence hypothesis. Materials and methods. To date, about 200 statistical criteria for testing the hypothesis of independence of a small sample are known. Using them together can significantly improve the quality of decisions made by the neural network. In this regard, the problem of effective comparative assessment of the achieved quality by one or another neural network arises. Results and conclusions. It is proposed to evaluate the quality of a neural network solution by the achieved increase in the sample when analyzing it using a single Pearson-Edleton-Edgeworth criterion of the late 19th century on a large sample in the most difficult situation of testing the independence hypothesis.
ISSN:2307-4205