Efficiency analysis using the machine learning algorithms: model development and verification

Recently, machine learning (ML) algorithms have been employed intensively in the field of finance as in all sectors. The issues such as financial distress prediction, bank credit risk calculation, etc., have been analyzed using ML algorithms. This study aimed to determine firm performance with the d...

Full description

Saved in:
Bibliographic Details
Published in:Quality & quantity Vol. 59; no. 4; pp. 3187 - 3209
Main Authors: Soydaş, Şafak Sönmez, Kalkan, Yusuf, Çam, Alper Veli, Barut, Abdulkadir
Format: Journal Article
Language:English
Published: Dordrecht Springer Netherlands 01.08.2025
Springer Nature B.V
Subjects:
Accuracy
Algorithms
Analysis
Automation
Banking
Banking industry
Business
Classification
Companies
Comparative analysis
Credit risk
Data
Data analysis
Data envelopment analysis
Datasets
Decision support systems
Efficiency
Financial analysis
Financial ratios
Lean manufacturing
Machine learning
Machinery
Methodology of the Social Sciences
Methods
Neural networks
Optimization
Organizational effectiveness
Organizational performance
Predictions
Psychological distress
Quality control
Raw materials
Sensitivity analysis
Social networks
Social Sciences
Software
Success
Supply chain management
Support vector machines
Tests
Verification
Logistic regression
Support vector machine
Business efficiency
Artificial neural networks
Data envelopment analysis
ISSN:0033-5177, 1573-7845
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recently, machine learning (ML) algorithms have been employed intensively in the field of finance as in all sectors. The issues such as financial distress prediction, bank credit risk calculation, etc., have been analyzed using ML algorithms. This study aimed to determine firm performance with the data envelopment analysis (DEA) method, sensitivity analysis, and ML algorithms and analyze the efficiency of companies via artificial neural networks (ANNs), support vector machines (SVMs), and logistic regression (LR) classification algorithms. In the study, first, 10 financial ratios were categorized into two parts, such as output and input, and efficiency scores were determined in MS Excel software. The obtained scores were included in the ML algorithm as a categorical dependent variable. Secondly, the data were extracted and included in the analysis software as 80% training and 20% test data, and the accuracy of ML algorithms was tested. Lastly, a comparative analysis of the estimation and classification algorithms of active and inactive companies was conducted. As a result of the analysis, the best classification prediction was seen as the ANN algorithm. SVM and LR algorithms also made an acceptable level of classification prediction. It was expected that the study would have contributed to the literature in terms of testing the companies whose efficiency scores were determined by the DEA method with ML techniques and determining which technique was more successful.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0033-5177
1573-7845
DOI:10.1007/s11135-025-02114-w