Page 1 of 9 - Cover Page Submission ID trn:oid:::1:3295499491

Rekha Puranik M
Employee Attrition Prediction Using Machine Learning
NEW

JSSATE STAFF

D M KUMAR

Document Details

Submission ID

trn:oid:::1:3295499491 5 Pages

Submission Date 3,090 Words

Jul 14, 2025, 12:00 PM GMT+5:30

18,780 Characters

Download Date

Jul 14, 2025, 12:02 PM GMT+5:30

File Name

conference_paper_2.docx

File Size

611.7 KB

Page 1 of 9 - Cover Page Submission ID trn:oid:::1:3295499491

 Page 2 of 9 - Integrity Overview Submission ID trn:oid:::1:3295499491

29% Overall Similarity

The combined total of all matches, including overlapping sources, for each database.

Filtered from the Report

Bibliography

Match Groups Top Sources

32 Not Cited or Quoted 29% 26% Internet sources

Matches with neither in-text citation nor quotation marks
6% Publications
0 Missing Quotations 0% 3% Submitted works (Student Papers)
Matches that are still very similar to source material

0 Missing Citation 0%
Matches that have quotation marks, but no in-text citation

0 Cited and Quoted 0%

Matches with in-text citation present, but no quotation marks

Integrity Flags
0 Integrity Flags for Review
Our system's algorithms look deeply at a document for any inconsistencies that
No suspicious text manipulations found. would set it apart from a normal submission. If we notice something strange, we flag
it for you to review.

A Flag is not necessarily an indicator of a problem. However, we'd recommend you

focus your attention there for further review.

Page 2 of 9 - Integrity Overview Submission ID trn:oid:::1:3295499491

 Page 3 of 9 - Integrity Overview Submission ID trn:oid:::1:3295499491

Match Groups Top Sources

32 Not Cited or Quoted 29% 26% Internet sources

Matches with neither in-text citation nor quotation marks
6% Publications
0 Missing Quotations 0% 3% Submitted works (Student Papers)
Matches that are still very similar to source material

0 Missing Citation 0%
Matches that have quotation marks, but no in-text citation

0 Cited and Quoted 0%

Matches with in-text citation present, but no quotation marks

Top Sources
The sources with the highest number of matches within the submission. Overlapping sources will not be displayed.

1 Internet

tnsroindia.org.in 20%

2 Internet

github.com 2%

3 Internet

www.coursehero.com 1%

4 Student papers

New College of the Humanities <1%

5 Publication

"Advances in Artificial-Business Analytics and Quantum Machine Learning", Sprin… <1%

6 Student papers

University of Brighton <1%

7 Student papers

universititeknologimara <1%

8 Student papers

VIT University <1%

9 Internet

icbsii.in <1%

10 Internet

www.codewithc.com <1%

Page 3 of 9 - Integrity Overview Submission ID trn:oid:::1:3295499491

 Page 4 of 9 - Integrity Overview Submission ID trn:oid:::1:3295499491

11 Internet

www.mdpi.com <1%

12 Publication

Inam Ullah Khan, Mariya Ouaissa, Mariyam Ouaissa, Muhammad Fayaz, Rehmat … <1%

13 Internet

inass.org <1%

14 Publication

Parneeta Dhaliwal, Manpreet Kaur, Hardeo Kumar Thakur, Rajeev Kumar Arya, Jo… <1%

15 Publication

Pawan Singh Mehra, Dhirendra Kumar Shukla. "Artificial Intelligence, Blockchain,… <1%

Page 4 of 9 - Integrity Overview Submission ID trn:oid:::1:3295499491

 Page 5 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491
5
Employee Attrition Prediction Using Machine
Learning
Ashalatha N S Bhumika P
3 Dept. of Information Science and Dept. of Information Science and
Engineering Engineering
JSS Academy Of Technical Education, JSS Academy Of Technical Education,
Bengaluru - 560060, INDIA Bengaluru - 560060, INDIA
ashalatha94@gmail.com Bhumikaprakash20@gmail.com

Abstract—Employee attrition is still a major problem in Logistic Regression and Multi-layer Perceptrons
organizational management as high turnover has an adverse (MLP) have high levels of prediction accuracy. [2-3].
impact on productivity, morale, and costs. In this paper,
6 supervised type of machine learning algorithms are used to However, issues regarding data imbalance (i.e., the
predict the employee attrition using the IBM HR Analytics majority of employees are not leaving), a high number
dataset with nine machine learning models tested both with and of attributes, and a need to make models interpretable
without dimensionality reduction and sampling techniques, still remain. Recent research has found solutions to
where Random Forest model accuracy was highest at 99.2%
when oversampling and Principal Component Analysis (PCA) is
those issues through the use of sampling (i.e., over-
applied. Feature importance is analyzed, and better approaches sampling and under-sampling techniques), and
for mitigating attrition are proposed. combined dimensional reduction (i.e., principal
component analysis - PCA) and ensemble learning
methods. [8-11]. The aim of this study is to build on
Keywords—Employee attrition, machine learning, HR that research by implementing and comparing multiple
10 analytics, random forest, PCA, oversampling, talent retention. ML classifiers Logistic Regression, Random Forest,
Support Vector Machines (SVM), Neural Networks
(MLP), Linear Discriminant Analysis (LDA) and other
I. INTRODUCTION ensemble learning to tackle data imbalance issues.

Employee attrition, or employees leaving a company

II. RELATED WORK
voluntarily or involuntarily, has become an
increasingly important problem for modern
organizations in the past few years. It can be disruptive The machine learning is used to predict employee
to workflow, lower workplace morale, increase attrition offers significant potential for enhancing
recruitment costs, increase training costs, and in some organizational strategies and minimizing human
cases, lead to the loss of institutional knowledge and resource expenses. Numerous studies have explored
core competencies. In a competitive, data-driven this domain, applying a wide array of models, datasets,
environment, employee retention has become more and analytics techniques with the shared goal of
strategic for organizations. [7][11]. Given the current identifying the most relevant features and developing
trend in the workplace of technology taking over roles robust predictive systems [1-4]. Recent studies have
previously occupied by employees both in terms of increasingly focused on ensemble learning to improve
data analytics and artificial intelligence companies the accuracy of prediction, rather than relying on a
have moved away from traditional heuristics utilized single model, ensemble methods combine multiple
by human resources (HR) departments and have classifiers such as Random Forest, Gradient Boosting,
utilized machine learning (ML)[8][11-12]. ML models and AdaBoost to leverage their complementary
are able to learn the patterns of terms of possible strengths. For example, researchers have proposed
attrition based on the HR data. This allows companies stacked ensemble frameworks with Logistic
to apply informed and, timely, personalized retention Regression as a meta-classifier, achieving significant
strategies aimed at reducing turnover rates. improvements in recall and F1 scores by aggregating
The attrition dataset has several variables which the strengths of base learners [6-9]. Beyond algorithm
describes the attrition behavior. The dataset contains choice, data preprocessing has emerged as a critical
attributes regarding age, income, overtime, role, step in model development. Methods such as missing
education, and work-life balance. Previous research value imputation, label encoding, and SMOTE
14 (Synthetic Minority Over-sampling Technique) for
exploring attrition behavior with this dataset has
7 indicated that supervised learning algorithms such as class balancing have demonstrated substantial impacts
Random Forest, Support Vector Machines (SVM), on model performance. In some cases, models saw

9 XXX-X-XXXX-XXXX-X/XX/$XX.00 ©20XX IEEE

Page 5 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491
 Page 6 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

marked improvements after applying these techniques methods help make predictive models transparent,
[2], [5]. A compelling example is presented by Ma et allowing organizations to understand why an
al.[21], who introduced a feature selection employee might leave, not just who might leave [22],
optimization framework integrating filter, wrapper, [24]. Feature selection remains a crucial step in
and embedded methods. This approach reduced aligning data science with HR strategy. Across many
dimensionality by 60% and, when paired with studies including those by Ma et al. [21] and Nandal et
XGBoost and SMOTE, achieved an accuracy of al. [23] variables such as jobSatisfaction, overtime,
87.3%. Importantly, the study used SHAP values to stockOptionLevel, workLifeBalance, and
interpret model predictions, identifying overtime, careerProgression have emerged as consistently strong
monthly income, and job involvement as key predictors. Selecting and emphasizing these features
predictors findings that align closely with HR domain enhances not only prediction performance but also the
knowledge. Similarly, Arqawi et al. [22] compared practical utility of models in decision-making.
ML models with deep learning architectures for Additionally, recent research is moving beyond
attrition prediction. Using the same dataset, their prediction toward prescriptive analytics. Scholars like
experiments showed that Multilayer Perceptrons Latorre et al. [23] have proposed benefit-driven
(MLP) outperformed other classifiers, reaching an F1 retention policies, while others have suggested
score of 94.5%. They also underscored the importance targeted interventions for high-risk groups based on
of data preprocessing, notably balancing the dataset role, demographics, or department. These
and encoding categorical variables. Their results developments emphasize translating model insights
confirm that well-tuned deep learning models, when into actionable HR strategies, reinforcing the value of
combined with appropriate data handling, can offer predictive analytics as a management tool.
high levels of predictive accuracy. Another
comparative study by Nandal et al. [23] evaluated a III. METHODOLOGY
wide spectrum of models including Logistic
Regression, KNN, Naive Bayes, ensemble methods
(e.g., AdaBoost, Random Forest), and deep learning
architectures like FNN and CNN. Their best-
performing model, FNN, achieved an accuracy of
97.5%. Their analysis emphasized job satisfaction,
overtime, and job involvement as strong correlates of
attrition.

In a different yet related context, Park et al. [24]

examined turnover intention a precursor to actual
attrition among new employees, especially recent
college graduates. Using a large-scale dataset of over
12,000 participants and ML models like logistic
regression, XGBoost, and KNN, they found that job
13 Figure 1: Architecture Diagram for Employee Attrition
security, career growth potential, and personal-job fit
Prediction
were the most influential predictors. Unlike studies
that rely solely on static employment data, this work
incorporated psychological and motivational factors, The proposed model is an end-to-end employee
highlighting a promising direction for improving attrition prediction workflow that automates predicting
12 employee churn through the application of multiple
prediction systems with behavioral data. Another
common theme across the literature is the integration machine learning algorithms. The architecture of the
of PCA with ensemble models, a hybrid approach that system includes steps for data collection,
has consistently outperformed standalone classifiers preprocessing, sampling, feature selection, training the
across diverse datasets [22], [23]. Such combinations model, evaluation, and interpretation.The workflow
strike a balance between computational efficiency and starts with collecting data from the HR dataset
predictive power, particularly in high-dimensional HR available on IBM, which is then processed by
datasets. Increasingly, the field is also shifting toward removing irrelevant attributes like EmployeeNumber
model interpretability and visualization. Techniques and StandardHours. All categorical values are encoded
like Self-Organizing Maps (SOM) have been used while numerical attributes are standardized via z-score
alongside decision trees to create cluster-based normalization. To address class imbalance, the system
visualizations that are more accessible to HR incorporates both random oversampling and
managers. Similarly, SHAP and other explainable AI undersampling. Dimensionality reduction is optional

Page 6 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

 Page 7 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

and is performed through Principal Component undersampled data with PCA but a decrease in the
Analysis (PCA), which enables the system to retain accuracy.
only the most meaningful features while improving
model performance. C. Decision Tree(DT)
15 The Decision Tree (DT) algorithm is a
supervised learning model to analyse learnings to
1 classify items into a class. The DT works by splitting
the data into smaller subsets, based on either Gini
impurity or information gain, until one

D. Random Forest (RF)

Figure 2. Project Methodology

This methodology includes training and evaluation on

nine different supervised machine learning classification
models. These models included straightforward ones
2 such as Logistic Regression, Naive Bayes, Decision
1
Tree, AdaBoost, Support Vector Machine, Linear
Figure 3. Feature Importance w.r.t. Random Forest
Discriminant Analysis, Multilayer Perceptron and K-
with Oversampling
Nearest Neighbors.
2 From Fig. 3, the most important features were
These models were trained on six distinct datasets:
MonthlyIncome followed by OverTime and Age,
imbalanced, undersampled, oversampled, PCA,
while the least important features were
undersampled with PCA, and oversampled with PCA,
PerformanceRating, Gender and BusinessTravel as
and assessed their performance. To optimize the
observed.
performance, hyperparameter tuning through
RandomSearchCV and GridSearchCV is used. E. AdaBoost
The best hyperparameters were the
1 Furthermore, K-fold cross-validation with 5 folds were
1 learningrate is set to 1.0 and n estimators set to 1000.
implemented on the training set. To address model It is also the second best performing model with high
interpretability, suitable graphs and figures are utilized. accuracy, precision and F1 score values.
The attrition decision is a biased metric, so the
11 evaluation of the models is done using a comprehensive F. Support Vector Machine (SVM)
1
set of metrics: accuracy, precision, recall, and F1 score. The model trained on over sampled data
performed the best and the model trained on the
1 A. Logistic Regression(LR)
imbalanced data performed the worst.
The Logistic Regression model performed the
best for oversampling with PCA, the hyperparamters
G. Linear Discriminant Analysis (LDA)
were: C = 0.1, penalty = L2 and solver = liblinear. The
The performance of LDA model is for
logistic regression model, admitted standardised data,
oversampled data with PCA with hyperparameters:
1 so it had higher accuracy.
shrinkage as auto and solver as ’lsqr’. LDA had higher
accuracy and precision for imbalanced data but higher
B. Naive Bayes(NB)
recall and F1 scores were observed for balanced data.
The Gaussian Naive Bayes model exhibited
the best recall for imbalanced and undersampled data,
H. K-Nearest Neighbors (KNN)
58.6% and 77.6% respectively. The oversampled and
1 For oversampling with PCA and had
1 undersampled data with PCA exhibited an increase in
hyperparameters: leaf size as 1, number of neighbors
precision, recall and F1 scores in oversampled and

Page 7 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

 Page 8 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

as 17 and weights as distance, the KNN model TABLE I. CLASSIFICATION SCORES FOR IMBALANCED DATA

performed well.
Model Accuracy Precision F1score
1 This was the only unsupervised model and had the
1 highest recall as well as good accuracy, precision, F1 LR 0.875 0.753 0.472
Score. NB 0.787 0.394 0.471
DT 0.836 0.498 0.351
I. MultiLayer Perceptron (MLP) RF 0.862 0.841 0.296
With hyperparameters: logistic activation
AdaBoost 0.858 0.769 0.293
function, alpha as 0.05 and lbfs solver, the best
performance with highest metrics was obtained for SVM 0.867 0.741 0.406
oversampled data with PCA followed by imbalanced LDA 0.867 0.683 0.439
and undersampled data. KNN 0.848 0.771 0.157
MLP 0.866 0.677 0.447
IV. EXPERIMENTAL SETUP AND RESULTS

TABLE II. CLASSIFICATION SCORES FOR UNDERSAMPLED DATA

Environment: The models were implemented using WITH PCA
Python 3.11 in Jupyter Notebook. Libraries used
included Scikit-learn, Pandas, Numpy, Matplotlib,
Seaborn, and imbalanced-learn. Model Accuracy Precision F1score
1 LR 0.758 0.751 0.761
4 Dataset: The IBM HR Analytics Employee Attrition
dataset is used. It consists of 1470 rows and 35 NB 0.680 0.649 0.709
columns. DT 0.841 0.796 0.852
RF 0.992 0.986 0.992
AdaBoost 0.989 0.982 0.990
SVM 0.951 0.912 0.953
LDA 0.609 0.610 0.667
KNN 0.894 0.826 0.904
MLP 0.946 0.905 0.949

1 TABLE III. CLASSIFICATION SCORES FOR OVERSAMPLED DATA

WITH PCA

Model Accuracy Precision F1score

1 LR 0.758 0.751 0.761
Figure 5. Employee Attrition Distribution NB 0.680 0.649 0.709
DT 0.841 0.796 0.852
RF 0.992 0.986 0.992
The variables "Yes" and "No" represent the target value, AdaBoost 0.989 0.982 0.990
attrition, which is a category variable. As seen in Fig. 5,
the dataset is extremely unbalanced and has substantially SVM 0.951 0.912 0.953
more samples for "No" than "Yes." LDA 0.609 0.610 0.667
KNN 0.894 0.826 0.904
After the further steps like preprocessing, converting MLP 0.946 0.905 0.949
categorical to numerical values, Oversampling and
Undersampling, Feature scaling and Principal
Component Analysis (PCA), the following results are For the imbalanced data, the logistic regression
obtained. model achieved best (87.5%) classification accuracy.
The below mentioned tables 1, 2 and 3 summarizes
1 the results of several classification evaluation metrics In the undersampled data with PCA, Random
for each and every model with Balanced/ Imbalanced Forest model accomplished best metrics with 72.4 %
data. accuracy and F1 score and 72.6 % precision and recall.

Page 8 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

 Page 9 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491

For the oversampled PCA data, tree based models [9] [9] M. S. Gazi, M. Nasiruddin, S. Dutta, R. Sikder, C. B. Huda, and M.
Z. Islam, “Employee attrition prediction in the USA: A machine
showed the best performance, of which Random Forest learning approach for HR analytics and talent retention strategies,” J.
showed the highest accuracy and F1-score, 99.2% and Bus. Manag. Stud., vol. 6, no. 3, pp. 47–54, May 2024. DOI:
10.32996/jbms.2024.6.3.6.
precision is 98.6%.
[10] [10] M. A. Akasheh, E. F. Malik, O. Hujran, and N. Zaki, “A decade
of research on machine learning techniques for predicting employee
5 V. CONCLUSION AND FUTURE WORKS turnover: A systematic literature review,” Expert Syst. Appl., vol. 238,
p. 121794, 2024. DOI: 10.1016/j.eswa.2023.121794.
[11] [11] A. V. Pachghare, S. Deshmukh, and S. Salunkhe, “Employee
8 This research presented a comprehensive machine churn prediction using machine learning techniques: A systematic
learning approach to predict employee attrition using review,” in Proc. 2024 Int. Conf. Data Sci. Netw. Security (ICDSNS),
2024, pp. 1–6. DOI: 10.1109/ICDSNS62112.2024.10691112..
the IBM HR Analytics dataset. Nine supervised [12] [12]H. Alqahtani, H. Almagrabi, and A. Alharbi, “Employee Attrition
models were evaluated, with Random Forest on Prediction Using Machine Learning Models: A Review Paper,” Int. J.
oversampled data with PCA achieving the highest of Artificial Intelligence and Applications (IJAIA), vol. 15, no. 2, pp.
24–43, Mar. 2024, DOI: 10.5121/ijaia.2024.15202
performance (accuracy: 99.2%, F1 score: 99.2%).
[13] [13] P. Latorre, H. López-Ospina, S. Maldonado, C. A. Guevara, and
Exploratory analysis revealed key attrition indicators J. Pérez, “Designing Employee Benefits to Optimize Turnover: A
such as age, overtime, and monthly income. Prescriptive Analytics Approach,” Computers & Industrial
Engineering, vol. 197, Art. 110582, 2024. DOI:
10.1016/j.cie.2024.110582K.
For future work, model deployment through web [14] [14] Degtyareva, D. A. Ageev, and V. V. Kukartsev, “Finding patterns
applications could enable real-time predictions for HR in employee attrition rates using self-organizing Kohonen maps and
decision trees,” in Proc. 2023 Int. Conf. Innovative Comput., Intell.
teams. Incorporating time-series data and external Commun. Smart Elect. Syst. (ICSESES), 2023, pp. 1–6. DOI:
economic indicators can improve prediction accuracy. 10.1109/ICSESES60034.2023.10465548.
Techniques like SHAP and LIME can be used for [15] [15] Doohee Chung et al., "Predictive model of employee attrition
based on stacking ensemble learning," Journal of Computer and
model interpretability. Additionally, adapting deep Information Sciences, vol. 20, no. 5, pp. 100-107, 2023.
learning architectures and evaluating cross-company [16] [16] Sonam Mittal, Ankita.,"Employee Attrition Prediction Using
generalizability would enhance robustness and real- Machine Learning Algorithms,"
Proc. 3rd Int. Conf. on Smart Generation Computing, Communication
world applicability. and Networking (SMART GENCON),
pp. 1–4, IEEE, Dec. 29–31, 2023, DOI:
REFERENCES 10.1109/SMARTGENCON60755.2023.10442776
[17] [17] F. Guerranti and G. M. Dimitri, “A Comparison of Machine
[1] [1] D.V. Lokeswar Reddy, Shake Hussain Basha, Vineeta Kaur Saluja, Learning Approaches for Predicting Employee Attrition,” Applied
Vandana Tiwari, Sreelekha T. K., and Aarti Sharma,"Optimizing Sciences, vol. 13, no. 1, Art. 267, pp. 1–8, 2023, DOI:
Employee Selection in Human Resource Management Using Decision 10.3390/app13010267
Tree Algorithms,"Proc. 6th Int. Conf. on Mobile Computing and
Sustainable Informatics (ICMCSI), IEEE, pp. 1857–1862, 2025, DOI: [18] [18] F. K. Alsheref, I. E. Fattoh, and W. M. Ead, "Automated Prediction
10.1109/ICMCSI64620.2025.10883407 of Employee Attrition Using Ensemble Model Based on Machine
Learning Algorithms," Computational Intelligence and Neuroscience,
[2] [2] M. Garba, M. Usman, and M. Saidu, “Enhancing employee attrition vol. 2022, Article ID 7728668, pp. 1–9, 2022, DOI:
prediction: The impact of data preprocessing on machine learning 10.1155/2022/7728668
model performance,” FUDMA J. Sci., vol. 9, no. 1, pp. 205–210, Jan.
2025. DOI: 10.33003/fjs-2025-0901-3030. [19] [19] D. Avrahami, D. Pessach, G. Singer, and H. Chalutz Ben-Gal, "A
Human Resources Analytics and Machine-Learning Examination of
[3] [3] Rajkumar Govindarajan, N. Komal Kumar, Sudhakar Reddy P., Sai Turnover: Implications for Theory and Practice," International Journal
Pravallika E., Dhatri B., and Pavan Kumar G.,"Predicting Employee of Manpower, vol. 43, no. 6, pp. 1405–1424, 2022, DOI: 10.1108/IJM-
Attrition: A Comparative Analysis of Machine Learning Models Using 12-2020-0548
the IBM Human Resource Analytics Dataset,"Procedia Computer
Science, vol. 258, pp. 4084–4093, 2025, DOI: [20] [20] L. Geiler, S. Affeldt, and M. Nadif, “An Effective Strategy for
10.1016/j.procs.2025.04.659 Churn Prediction and Customer Profiling,” Data & Knowledge
Engineering, vol. 142, Art. 102100, 2022, DOI:
[4] [4]Nishitha Reddy Nalla. "Machine Learning Models for Predicting 10.1016/j.datak.2022.102100
Employee Retention and Performance,"
International Journal of Data Science and Machine Learning [21] [21] D. Ma, M. Shu, and H. Zhang, "Feature Selection Optimization
(IJDSML), pp. 15–19, Vol. 05, Issue 01, Feb. 2025, DOI: for Employee Retention Prediction: A Machine Learning Approach,"
10.55640/ijdsml-05-01-04 Preprints, pp. 1–11, Apr. 2025, DOI:
10.20944/preprints202504.1549.v1.
[5] [5]A. Benabou, F. Touhami, and M.A. Sabri, "Predicting Employee
Turnover Using Machine Learning Techniques," Acta Informatica [22] [22] S. M. Arqawi, M. A. Abu Rumman, E. A. Zitawi, A. H. Rabaya,
Pragensia, vol. 14, no. 1, pp. 112–127, 2025, DOI: 10.18267/j.aip.255 A. S. Sadaqa, B. S. Abunasser, and S. S. Abu-Naser, "Predicting
Employee Attrition and Performance Using Deep Learning," Journal
[6] [6] R. Govindarajan, N. K. Kumar, S. P. Reddy, S. E. Pravallika, B. of Theoretical and Applied Information Technology, vol. 100, no. 21,
Dhatri, and P. G. Kumar, "Predicting Employee Attrition: A pp. 6526–6533, Nov. 2022.
Comparative Analysis of Machine Learning Models Using the IBM
Human Resource Analytics Dataset," Procedia Computer Science, vol. [23] [23] M. Nandal, V. Grover, D. Sahu, and M. Dogra, "Employee
258, pp. 4084–4093, 2025, doi: 10.1016/j.procs.2025.04.659. Attrition: Analysis of Data Driven Models," EAI Endorsed
Transactions on Internet of Things, vol. 10, no. 2, pp. 1–7, Jan. 2024,
[7] [7]P. Nagpal, A. Pawar, and S. H. M., “Predicting employee attrition DOI: 10.4108/eetiot.4762.
through HR analytics: A machine learning approach,” in Proc. 4th Int.
Conf. Innovative Practices Technol. Manag. (ICIPTM), 2024, pp. 1–6. [24] [24] J. Park, Y. Feng, and S.-P. Jeong, "Developing an Advanced
DOI: 10.1109/ICIPTM59628.2024.10563285. Prediction Model for New Employee Turnover Intention Utilizing
Machine Learning Techniques," Scientific Reports, vol. 14, no. 1221,
[8] [8] Aitong Jin et al., "Predicting Employee Attrition Using Machine pp. 1–15, 2024, DOI: 10.1038/s41598-023-50593-4.
Learning Approaches," International Journal of Computer
Applications, vol. 182, no. 45, pp. 12-19, 2024.

Page 9 of 9 - Integrity Submission Submission ID trn:oid:::1:3295499491