DFCEA: Deep Forest Classifier-Based Employee Attrition Prediction

Mahmut Tokmak

Makale detayları

DFCEA: Deep Forest Classifier-Based Employee Attrition Prediction

Yazarlar:

Mahmut Tokmak

Yayıncı:

Istanbul University

Ana Başlıklar:

Makine Öğrenme (Diğer)

Dergi:

Acta Infologica Vol.9 Issue 1, pp.19-33

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001354871

pISSN:

2602-3563

Anahtar kelimeler:

Özet:

 For companies, employee attrition is an important issue because human resources are the most important resources of a company. In companies, employee attrition can have different causes. However, human resource managers must recognize employee attrition indicators in the early stages. Employee attrition can lead to organizational losses for various reasons, such as interruption of work, interruption of tasks that need to be performed, the cost of re-employment and retraining, and the risk of information leakage. Therefore, in this study, DFCEA: Deep Forest Classifier-Based Employee Attrition prediction model is proposed to predict employee attrition. Thus, this study aimed to help company managers take measures to prevent the loss of human resources. The IBM HR Analytics Employee Attrition  Performance dataset was used in this study. The dataset was subjected to data cleaning, data encoding, data normalization, and data balancing preprocessing. The model was then trained and tested using the Deep Forest algorithm. With the proposed method, 98.8 accuracy and 98.8 f1 score were obtained. The obtained performance metrics are compared with known machine learning methods and other studies, and the performance power of the proposed method is demonstrated. The results demonstrate that the proposed DFCEA framework is highly effective in predicting employee attrition. Therefore, the framework presented in this study can help researchers, organization leaders, and human resource professionals predict employee attrition and contribute to the development of new prediction models.

 For companies, employee attrition is an important issue because human resources are the most important resources of a company. In companies, employee attrition can have different causes. However, human resource managers must recognize employee attrition indicators in the early stages. Employee attrition can lead to organizational losses for various reasons, such as interruption of work, interruption of tasks that need to be performed, the cost of re-employment and retraining, and the risk of information leakage. Therefore, in this study, DFCEA: Deep Forest Classifier-Based Employee Attrition prediction model is proposed to predict employee attrition. Thus, this study aimed to help company managers take measures to prevent the loss of human resources. The IBM HR Analytics Employee Attrition  Performance dataset was used in this study. The dataset was subjected to data cleaning, data encoding, data normalization, and data balancing preprocessing. The model was then trained and tested using the Deep Forest algorithm. With the proposed method, 98.8 accuracy and 98.8 f1 score were obtained. The obtained performance metrics are compared with known machine learning methods and other studies, and the performance power of the proposed method is demonstrated. The results demonstrate that the proposed DFCEA framework is highly effective in predicting employee attrition. Therefore, the framework presented in this study can help researchers, organization leaders, and human resource professionals predict employee attrition and contribute to the development of new prediction models.

 For companies, employee attrition is an important issue because human resources are the most important resources of a company. In companies, employee attrition can have different causes. However, human resource managers must recognize employee attrition indicators in the early stages. Employee attrition can lead to organizational losses for various reasons, such as interruption of work, interruption of tasks that need to be performed, the cost of re-employment and retraining, and the risk of information leakage. Therefore, in this study, DFCEA: Deep Forest Classifier-Based Employee Attrition prediction model is proposed to predict employee attrition. Thus, this study aimed to help company managers take measures to prevent the loss of human resources. The IBM HR Analytics Employee Attrition  Performance dataset was used in this study. The dataset was subjected to data cleaning, data encoding, data normalization, and data balancing preprocessing. The model was then trained and tested using the Deep Forest algorithm. With the proposed method, 98.8 accuracy and 98.8 f1 score were obtained. The obtained performance metrics are compared with known machine learning methods and other studies, and the performance power of the proposed method is demonstrated. The results demonstrate that the proposed DFCEA framework is highly effective in predicting employee attrition. Therefore, the framework presented in this study can help researchers, organization leaders, and human resource professionals predict employee attrition and contribute to the development of new prediction models.

 For companies, employee attrition is an important issue because human resources are the most important resources of a company. In companies, employee attrition can have different causes. However, human resource managers must recognize employee attrition indicators in the early stages. Employee attrition can lead to organizational losses for various reasons, such as interruption of work, interruption of tasks that need to be performed, the cost of re-employment and retraining, and the risk of information leakage. Therefore, in this study, DFCEA: Deep Forest Classifier-Based Employee Attrition prediction model is proposed to predict employee attrition. Thus, this study aimed to help company managers take measures to prevent the loss of human resources. The IBM HR Analytics Employee Attrition  Performance dataset was used in this study. The dataset was subjected to data cleaning, data encoding, data normalization, and data balancing preprocessing. The model was then trained and tested using the Deep Forest algorithm. With the proposed method, 98.8 accuracy and 98.8 f1 score were obtained. The obtained performance metrics are compared with known machine learning methods and other studies, and the performance power of the proposed method is demonstrated. The results demonstrate that the proposed DFCEA framework is highly effective in predicting employee attrition. Therefore, the framework presented in this study can help researchers, organization leaders, and human resource professionals predict employee attrition and contribute to the development of new prediction models.

Makale detayları

DFCEA: Deep Forest Classifier-Based Employee Attrition Prediction

Yazarlar:

Mahmut Tokmak

Yayıncı:

Istanbul University

Ana Başlıklar:

Makine Öğrenme (Diğer)

Dergi:

Acta Infologica Vol.9 Issue 1, pp.19-33

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001354871

pISSN:

2602-3563

Artificial Intelligence

Machine Learning

Deep Forest

Employee Attrition

Sayı

Makale detayları

DFCEA: Deep Forest Classifier-Based Employee Attrition Prediction

Yazarlar: Mahmut Tokmak

Yayıncı: Istanbul University

Ana Başlıklar: Makine Öğrenme (Diğer)

Dergi: Acta Infologica Vol.9 Issue 1, pp.19-33

Belge Tipi: Makale

Yayın Tarihi: 6/2025

Belge Dili: İngilizce

Document ID (Org No.): 2001354871

pISSN: 2602-3563

Artificial Intelligence

Machine Learning

Deep Forest

Employee Attrition

Sayı

Yazarlar:

Mahmut Tokmak

Yayıncı:

Istanbul University

Ana Başlıklar:

Makine Öğrenme (Diğer)

Dergi:

Acta Infologica Vol.9 Issue 1, pp.19-33

Belge Tipi:

Makale

Yayın Tarihi:

6/2025

Belge Dili:

İngilizce

Document ID (Org No.):

2001354871

pISSN:

2602-3563