Gaussian process regression ́s hyperparameters optimization to predict financial distress
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F75081431%3A_____%2F23%3A00002637" target="_blank" >RIV/75081431:_____/23:00002637 - isvavai.cz</a>
Result on the web
<a href="https://retos.ups.edu.ec/index.php/retos/article/view/7417" target="_blank" >https://retos.ups.edu.ec/index.php/retos/article/view/7417</a>
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
Gaussian process regression ́s hyperparameters optimization to predict financial distress
Original language description
Predicting financial distress has become one of the most important topics of the hour that has swept the accounting and financial field due to its significant correlation with the development of science and technology. The main objective of this paper is to predict financial distress based on the Gaussian Process Regression (GPR) and then compare the results of this model with the results of other deep learning models (SVM, LR, LD, DT, KNN). The analysis is based on a dataset of 352 companies extracted from the Kaggle database. As for predictors, 83 financial ratios were used. The study concluded that the use of GPR achieves very relevant results. Furthermore, it outperformed the rest of the deep learning models and achieved first place equally with the SVM model with a classification accuracy of 81%. The results contribute to the maintenance of the integrated system and the prosperity of the country's economy, the prediction of the financial distress of companies and thus the potential prevention of disruption of the given system.
Czech name
—
Czech description
—
Classification
Type
J<sub>imp</sub> - Article in a specialist periodical, which is included in the Web of Science database
CEP classification
—
OECD FORD branch
50200 - Economics and Business
Result continuities
Project
—
Continuities
V - Vyzkumna aktivita podporovana z jinych verejnych zdroju
Others
Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Name of the periodical
RETOS-REVISTA DE CIENCIAS DE LA ADMINISTRACION Y ECONOMIA
ISSN
1390-6291
e-ISSN
—
Volume of the periodical
13
Issue of the periodical within the volume
26
Country of publishing house
ES - SPAIN
Number of pages
17
Pages from-to
273-289
UT code for WoS article
001077774200006
EID of the result in the Scopus database
—