Research on Financial Difficulties Prediction and Optimization Strategies of Small and Medium-sized Enterprises Based on Support Vector Machine

In order to improve the accuracy of financial distress prediction for small and medium-sized enterprises (SMEs) and enhance the prediction performance, this paper adopts principal component analysis (PCA) to extract the input variables required for financial distress prediction, and applies Gray Wol...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Applied mathematics and nonlinear sciences Ročník 10; číslo 1
Hlavný autor: Duan, Mingyue
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Beirut Sciendo 01.01.2025
De Gruyter Brill Sp. z o.o., Paradigm Publishing Services
Predmet:
97B20
Accuracy
Datasets
Financial distress prediction
Gray wolf optimization algorithm
Optimization
Principal component analysis
Small & medium sized enterprises-SME
Support vector machine
Support vector machines
ISSN:2444-8656, 2444-8656
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:In order to improve the accuracy of financial distress prediction for small and medium-sized enterprises (SMEs) and enhance the prediction performance, this paper adopts principal component analysis (PCA) to extract the input variables required for financial distress prediction, and applies Gray Wolf Optimization Algorithm (GWO) to the optimization of penalty coefficients and kernel function parameters of the Support Vector Machine model (SVM) to propose a financial distress prediction model based on GWO-SVM. The performance of this paper’s GWO-SVM model is evaluated in terms of fitness and prediction classification accuracy. On the same dataset, the CPU time of this paper’s model is 48.44s, the classification accuracy can reach 87.01%, and the classification accuracy in the confusion matrix results can reach up to 93.48%, which outperforms other grid search, GA, PSO, and GWO comparison models. In addition, the prediction accuracy of this paper’s model is always higher than that of the SVM and ICA-SVM models as comparisons, whether in the balanced sample dataset or the unbalanced sample dataset, and maintains the prediction accuracy level of more than 80% and 70% in different datasets, respectively.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2444-8656
2444-8656
DOI:10.2478/amns-2025-1097