Predicting Employee Under Stress for Preemptive Remediation Using Machin Learning Algorithms

Abstract

The modern world is filled with stress. A person\'s pressure is affected by a variety of factors. Representatives in IT are more likely to be under pressure due to work pressure, overburdening, higher worker mastery, and so on. When a person is stressed, it can lead to a variety of mental health issues such as depression, anxiety, somatization, lack of concentration, andsoon.Asaresult,itisnecessarytoidentifyhumanstressatan earlystageinordertoprovide appropriate solutions and alleviate stress. There has been a lot of research done on stress prediction. Many research papers use Machine Learning techniques to predict stress, and many papersuse IOT -based sensors to extract the features needed for stress prediction. There are so many existing systems in predicting employee stress. In this project we are going to predict the employee stress using XGBOOST algorithm since it gives more Accuracy. Based on this the prediction we will remediate the persons under stress in the early stage which is good for their health and as well as the work.

Introduction

The COVID-19 pandemic, declared by WHO in March 2020, caused widespread disruption globally, affecting economies and increasing stress among employees due to prolonged uncertainty and heavy workloads. This research aims to analyze employee stress levels during the pandemic using machine learning (ML) techniques to predict stress without intrusive hardware.

Traditional methods of stress assessment are often subjective and limited, whereas ML models can analyze diverse data (biometric signals, workplace behavior, communication) to predict stress accurately. Common ML algorithms used include Logistic Regression, SVM, Random Forest, and deep learning models.

Challenges include data privacy, difficulties in collecting accurate data, imbalanced datasets, individual variability in stress triggers, and the complexity of real-time prediction and model interpretability.

The proposed methodology uses the XGBoost algorithm to analyze anonymized data from multiple sources (digital behavior, workplace logs, communication) without requiring external sensors. It offers high accuracy, early stress detection, handles mixed data types, and provides feature importance insights for HR teams.

Different ML types are discussed: supervised (uses labeled data), unsupervised (finds patterns in unlabeled data), and semi-supervised learning (combines both). Applications of ML in speech recognition, customer service chatbots, computer vision, recommendation engines, and automated stock trading are also highlighted.

The system architecture includes data preprocessing (handling missing data, encoding categorical variables, normalization), splitting data into training/testing sets, training the XGBoost model, and evaluating it using multiple metrics such as accuracy and F1-score.

Conclusion

To evaluate our model to achieve a better performance which is done by using XGB classifier. This is one of the best optimization technique and this is like a decision tree-based algorithm which adopts gradient boosting frame work technique for analysis and confusion matrix which tells us how many correct values are predicted by our model. XG Boost has tremendous predictive power and is about 10 times more durable than other gradient boosting techniques. It holds a varietyof regularization which diminishes overfitting and enhances overall performance. Consequently, it is further recognized as the “regularized boosting” technique. Like it has true positive, true negative, false positive, false negative values. Used to evaluate the performance of the classification model

References

[1] Shekhar Pandey, Supriya Muthuraman, Abhilash Shrivastava.The International Symposium on Intelligent Systems Technologies and Applications (2018), DOI: 10.1007/978-3-319- 68385- 0_10. [2] Ramachandran, R; Rajeev, D.C; Krishnan, S.G; Subathra.P. International Journal of Applied Engineering Research (2015), Research India Publications, Volume 10,Number 10, p.25433- 25448 [3] RaminZibaseresht: How to Respond to the Ongoing Pandemic Outbreak of the Coronavirus Disease (COVID-19) (WHO- World Health Organization) (2020), ISSN 2349- 8870. [4] Chen, Tianqi; Guestrin, Carlos; “XG Boost: A scalable Tree Boosting System”. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and DataMining, San Francisco, USA (2016). ACM. pp. 785-794. [5] Jolliffe I.T. Principal Component Analysis, Series: Springer Series in Statistics, 2nd ed., Springer, NY, 2002,XXIX,487p.28 illus. ISBN 978-0-387-95442-4. [6] Manuela Aparicio and Carlos J. Costa “Data Visualization”. Communication Design Quarterly (2014). DOI: 10.1145/2721882.2721883. [7] Ali.M., Alqahtani.A.,Jones.M.W., Xie.X .“Clustering and classification for Time Series Data in Visual Analytics: A Survey IEEE Access 7,8930535, pp. 181314-181338. [8] Mouubayedd.A, Injadat.M.,Nassif, Lutfuyya, H.Shami, A E-learning: Challenges and Research Opportunities Using Machine Learning and Data Analytics (2018) IEEE Access 6,8417405. pp. 39117-39138. [9] Kim., Soyata, T., Behnagh, R.F. Towards Emotionally Aware AI Smart Classroom: Current Issues and Directions for Engineering and Education (2018) IEEE pp. 5308-5331. [10] PriyonesiS.Madeh; EI-Diraby Tamer E. “Role of Data Analytics in Infrastructure Asset Management: Overcoming Data Size and Quality Problems”. Journal of Transprotation Engineering, Pavements (2020). DOI: 10.1061/JPEODX.0000175. [11] AbhijeetRawal, SnehaMhatre.IOSRJournal ofBussiness and Management (IOSR-JBM), e-ISSN:2278-487X,P-ISSN:2319 [12] JanneSkakon, Karina Nielsen, Vihelm Borg, Jaime Gazman. An international Journal of work, Health and organizations, Volume 24,2010-Issue 2. [13] K. S. Santosh and S. H. Bharathi, \"Non-negative matrix factorization algorithms for blind source sepertion in speech recognition,\" 2017 2nd IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT), Bangalore, 2017,pp.2242-2246,doi: 10.1109/RTEICT.2017.8256999. [14] Rajendra Prasad P, N. Narayan, S. Gayathri and S. Ganna, “An Efficient E-Health Monitoring with Smart Dispensing System for Remote Areas”, 2018 3rd IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT), Bangalore, India, 2018, pp. 2120-2124. doi: 10.1109/RTEICT42901.2018.9012480

Copyright

Copyright © 2025 K.Tulasi Krishna Kumar, N. Dharmaraju Reddy. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.