Federated Learning for Privacy-Preserving Employee Performance Analytics
With the increasing sensitivity surrounding employee performance data, there is a pressing need for predictive systems that preserve privacy while delivering actionable insights to organizations. This paper introduces HFAN-Priv, a hierarchical federated attention network designed to predict employee...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11087544/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839608043862491136 |
|---|---|
| author | Jay Barach |
| author_facet | Jay Barach |
| author_sort | Jay Barach |
| collection | DOAJ |
| description | With the increasing sensitivity surrounding employee performance data, there is a pressing need for predictive systems that preserve privacy while delivering actionable insights to organizations. This paper introduces HFAN-Priv, a hierarchical federated attention network designed to predict employee resignation risk and evaluate performance trends without sharing raw data across organizations. The framework integrates feature-level and instance-level attention to model complex workforce patterns, applies differential privacy through gradient masking to ensure compliance with data protection regulations, and enhances interpretability using local SHAP and LIME explanations. Experiments conducted on a real-world employee productivity dataset show that HFAN-Priv achieves near-perfect predictive accuracy, robust cross-client generalization, and transparent decision-making, all while maintaining strong privacy guarantees. The proposed approach presents a scalable, ethical, and effective solution for HR analytics in decentralized environments. |
| format | Article |
| id | doaj-art-ff3a33a94e3349ca891e8782a0536d41 |
| institution | Matheson Library |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-ff3a33a94e3349ca891e8782a0536d412025-07-31T23:00:42ZengIEEEIEEE Access2169-35362025-01-011313272413273810.1109/ACCESS.2025.359136011087544Federated Learning for Privacy-Preserving Employee Performance AnalyticsJay Barach0https://orcid.org/0009-0009-0416-9712Systems Staffing Group, Inc., Norristown, PA, USAWith the increasing sensitivity surrounding employee performance data, there is a pressing need for predictive systems that preserve privacy while delivering actionable insights to organizations. This paper introduces HFAN-Priv, a hierarchical federated attention network designed to predict employee resignation risk and evaluate performance trends without sharing raw data across organizations. The framework integrates feature-level and instance-level attention to model complex workforce patterns, applies differential privacy through gradient masking to ensure compliance with data protection regulations, and enhances interpretability using local SHAP and LIME explanations. Experiments conducted on a real-world employee productivity dataset show that HFAN-Priv achieves near-perfect predictive accuracy, robust cross-client generalization, and transparent decision-making, all while maintaining strong privacy guarantees. The proposed approach presents a scalable, ethical, and effective solution for HR analytics in decentralized environments.https://ieeexplore.ieee.org/document/11087544/Federated learningemployee performanceprivacy-preserving AIdifferential privacySHAPhuman resource analytics |
| spellingShingle | Jay Barach Federated Learning for Privacy-Preserving Employee Performance Analytics IEEE Access Federated learning employee performance privacy-preserving AI differential privacy SHAP human resource analytics |
| title | Federated Learning for Privacy-Preserving Employee Performance Analytics |
| title_full | Federated Learning for Privacy-Preserving Employee Performance Analytics |
| title_fullStr | Federated Learning for Privacy-Preserving Employee Performance Analytics |
| title_full_unstemmed | Federated Learning for Privacy-Preserving Employee Performance Analytics |
| title_short | Federated Learning for Privacy-Preserving Employee Performance Analytics |
| title_sort | federated learning for privacy preserving employee performance analytics |
| topic | Federated learning employee performance privacy-preserving AI differential privacy SHAP human resource analytics |
| url | https://ieeexplore.ieee.org/document/11087544/ |
| work_keys_str_mv | AT jaybarach federatedlearningforprivacypreservingemployeeperformanceanalytics |