Deep generative neural network dynamics modeling and employee satisfaction prediction for human resource management

Jie Wang; Jie Wang

doi:10.3934/math.2026543

AIMS Mathematics

2026, Volume 11, Issue 5: 13174-13195. doi: 10.3934/math.2026543

Previous Article Next Article

Research article Special Issues

Deep generative neural network dynamics modeling and employee satisfaction prediction for human resource management

Jie Wang ^,

School of Accounting, Hangzhou College of Commerce, Zhejiang Gongshang University, Hangzhou 310000, China

Received: 06 February 2026 Revised: 29 March 2026 Accepted: 07 April 2026 Published: 12 May 2026
MSC : 34L99, 68T07

Traditional human resource management has limited ability to capture the temporal evolution of employee satisfaction and to learn informative representations from high-dimensional, multi-source data, thereby constraining the accuracy and interpretability of dynamic prediction. To address this issue, I developed a joint VAE-Neural ODE framework that integrates a variational autoencoder (VAE) with a neural ordinary differential equation (neural ODE). Specifically, the VAE encoder employs the reparameterization trick to learn the latent probability distribution of heterogeneous employee records and generate compact yet informative latent features. On this basis, the neural ODE models the continuous-time evolution of latent states and propagates them through a differentiable ODE solver for temporal forecasting. The decoder then maps the evolved latent state to a continuous employee satisfaction score normalized to the interval [0, 1]. Accordingly, the prediction task was formulated primarily as a regression problem, and model performance was evaluated using mean absolute error (MAE) and root mean squared error (RMSE). For interpretive comparison, the continuous outputs were further discretized using a fixed threshold to derive a binary accuracy metric. Experimental results showed that the proposed method achieves an accuracy of 94.7%, an MAE of 0.031, and an RMSE of 0.052 on the test set, while maintaining 87.2% accuracy in cross-quarter forecasting. These findings indicate that the proposed framework can effectively capture nonlinear satisfaction dynamics and provide quantitative support for data-driven human resource analysis. Moreover, by formulating employee satisfaction prediction as a latent continuous-time dynamical process and evolving hidden states through an ODE flow rather than purely discrete recursive updates, the framework produces smoother latent trajectories and reduces error accumulation over longer forecasting horizons and under irregular observation intervals, which are common in non-uniformly sampled HR records.
- human resource management,
- employee satisfaction prediction,
- variational autoencoder,
- neural ordinary differential equation,
- dynamic modeling
Citation: Jie Wang. Deep generative neural network dynamics modeling and employee satisfaction prediction for human resource management[J]. AIMS Mathematics, 2026, 11(5): 13174-13195. doi: 10.3934/math.2026543

Related Papers:

Abstract

Traditional human resource management has limited ability to capture the temporal evolution of employee satisfaction and to learn informative representations from high-dimensional, multi-source data, thereby constraining the accuracy and interpretability of dynamic prediction. To address this issue, I developed a joint VAE-Neural ODE framework that integrates a variational autoencoder (VAE) with a neural ordinary differential equation (neural ODE). Specifically, the VAE encoder employs the reparameterization trick to learn the latent probability distribution of heterogeneous employee records and generate compact yet informative latent features. On this basis, the neural ODE models the continuous-time evolution of latent states and propagates them through a differentiable ODE solver for temporal forecasting. The decoder then maps the evolved latent state to a continuous employee satisfaction score normalized to the interval [0, 1]. Accordingly, the prediction task was formulated primarily as a regression problem, and model performance was evaluated using mean absolute error (MAE) and root mean squared error (RMSE). For interpretive comparison, the continuous outputs were further discretized using a fixed threshold to derive a binary accuracy metric. Experimental results showed that the proposed method achieves an accuracy of 94.7%, an MAE of 0.031, and an RMSE of 0.052 on the test set, while maintaining 87.2% accuracy in cross-quarter forecasting. These findings indicate that the proposed framework can effectively capture nonlinear satisfaction dynamics and provide quantitative support for data-driven human resource analysis. Moreover, by formulating employee satisfaction prediction as a latent continuous-time dynamical process and evolving hidden states through an ODE flow rather than purely discrete recursive updates, the framework produces smoother latent trajectories and reduces error accumulation over longer forecasting horizons and under irregular observation intervals, which are common in non-uniformly sampled HR records.

References

[1]	C. H. Chang, H. W. Lin, W. H. Tsai, W. L. Wang, C. T. Huang, Employee satisfaction, corporate social responsibility and financial performance, Sustainability, 13 (2021), 9996. https://doi.org/10.3390/su13189996 doi: 10.3390/su13189996
[2]	B. Miethlich, M. Beliakova, L. Voropaeva, O. Ustyuzhina, T. Yurieva, Internal corporate policy: CSR and employee satisfaction, Employ. Respons. Rights J., 35 (2023), 127–141. https://doi.org/10.1007/s10672-022-09406-5 doi: 10.1007/s10672-022-09406-5
[3]	P. Ruiz-Palomino, R. Morales-Sánchez, R. Martínez-Cañas, Corporate sustainability, ethics and employee satisfaction, Sustainability, 13 (2021), 11964. https://doi.org/10.3390/su132111964 doi: 10.3390/su132111964
[4]	E. Uchida, Y. Kino, Study on the relationship between employee satisfaction and corporate performance in Japan via text mining, Procedia Comput. Sci., 192 (2021), 1730–1739. https://doi.org/10.1016/j.procs.2021.08.178 doi: 10.1016/j.procs.2021.08.178
[5]	T. V. D. Voordt, P. A. Jensen, The impact of healthy workplaces on employee satisfaction, productivity and costs, J. Corp. Real Estate, 25 (2023), 29–49. https://doi.org/10.1108/jcre-03-2021-0012 doi: 10.1108/jcre-03-2021-0012
[6]	T. Zhang, J. Zhang, S. Tu, An empirical study on corporate ESG behavior and employee satisfaction: A moderating mediation model, Behav. Sci., 14 (2024), 274. https://doi.org/10.3390/bs14040274 doi: 10.3390/bs14040274
[7]	I. N. Latifah, A. A. Suhendra, I. Mufidah, Factors affecting job satisfaction and employee performance: A case study in an Indonesian sharia property companies, Int. J. Prod. Perform. Manage., 73 (2024), 719–748. https://doi.org/10.1108/ijppm-03-2021-0132 doi: 10.1108/ijppm-03-2021-0132
[8]	S. Tabassum, S. Alam, M. S. Akter, S. Ahmad, Factors influencing employee satisfaction in workplace, Int. J. Eng. Appl. Sci. Technol., 7 (2023), 61–67. https://doi.org/10.33564/ijeast.2023.v07i12.010 doi: 10.33564/ijeast.2023.v07i12.010
[9]	J. Kim, P. S. Chang, S. B. Yang, I. Choi, B. Lee, A comparative analysis of job satisfaction prediction models using machine learning: A mixed-method approach, Data Technol. Appl., 59 (2025), 41–60. https://doi.org/10.1108/dta-10-2023-0697 doi: 10.1108/dta-10-2023-0697
[10]	M. E. Zhuang, W. T. Pan, Data modelling in human resource management: influencing factors of employees' job satisfaction, Math. Probl. Eng., 2022 (2022), 3588822. https://doi.org/10.1155/2022/3588822 doi: 10.1155/2022/3588822
[11]	P. Kang, Predicting HR management strategies and employee satisfaction in enterprises based on deep generative models, Appl. Math. Nonlinear Sci., 10 (2025), 1–16. https://doi.org/10.2478/amns-2025-0043 doi: 10.2478/amns-2025-0043
[12]	F. Rustam, I. Ashraf, R. Shafique, A. Mehmood, S. Ullah, G. S. Choi, Review prognosis system to predict employees job satisfaction using deep neural network, Comput. Intell., 37 (2021), 924–950. https://doi.org/10.1111/coin.12440 doi: 10.1111/coin.12440
[13]	J. S. Sekhon, H. Sadawarti, Prediction of employee satisfaction level using optimized neural network, Int. J. Adv. Res. Eng. Technol., 12 (2021), 129–143. https://doi.org/10.34218/IJARET.12.1.2021.011 doi: 10.34218/IJARET.12.1.2021.011
[14]	M. Y. M. Adib, S. Chakraborty, M. T. Waishy, M. H. K. Mehedi, A. A. Rasel, BiLSTM-ANN based employee job satisfaction analysis from glassdoor data using web scraping, Procedia Comput. Sci., 222 (2023), 1–10. https://doi.org/10.1016/j.procs.2023.08.139 doi: 10.1016/j.procs.2023.08.139
[15]	M. Qadir, I. Noreen, A. A. Shah, Bi-LSTM deep learning approach for employee churn prediction, J. Inf. Commun. Technol. Robot. Appl., 12 (2021), 1–10. https://doi.org/10.51239/jictra.v0i0.255 doi: 10.51239/jictra.v0i0.255
[16]	T. Kanchinadam, Z. Meng, J. Bockhorst, V. Singh, G. Fung, Graph neural networks to predict customer satisfaction following interactions with a corporate call center, arXiv: 2102.00420, 2021. https://doi.org/10.48550/arXiv.2102.00420
[17]	Z. Yue, M. R. A. Razak, L. Li, Dynamic context‐weighted embeddings: A novel approach to predictive modelling, IET Commun., 19 (2025), e70036. https://doi.org/10.1049/cmu2.70036 doi: 10.1049/cmu2.70036
[18]	X. Li, L. Zhang, D. Li, D. Guo, Construction and simulation of a strategic HR decision model based on recurrent neural network, J. Math., 2022 (2022), 5390176. https://doi.org/10.1155/2022/5390176 doi: 10.1155/2022/5390176
[19]	H. K. Sriram, Generative AI and neural networks in human resource management: Transforming payroll, workforce insights, and digital employee payments through AI innovations, Adv. Consum. Res., 2 (2025), 250–260.
[20]	H. Du, S. Du, W. Li, Probabilistic time series forecasting with deep non‐linear state space models, CAAI Trans. Intell. Technol., 8 (2023), 3–13. https://doi.org/10.1049/cit2.12085 doi: 10.1049/cit2.12085
[21]	H. Han, Z. Liu, M. B. Barrios, J. Li, Z. Zeng, N. Sarhan, et al., Time series forecasting model for non-stationary series pattern extraction using deep learning and GARCH modeling, J. Cloud Comp., 13 (2024), 2. https://doi.org/10.1186/s13677-023-00576-7 doi: 10.1186/s13677-023-00576-7
[22]	L. Zhang, R. Wang, Z. Li, J. Li, Y. Ge, S. Wa, Time-series neural network: a high-accuracy time-series forecasting method based on Kernel filter and time attention, Information, 14 (2023), 500. https://doi.org/10.3390/info14090500 doi: 10.3390/info14090500
[23]	M. A. Khan, S. A. Iqbal, M. S. Khan, M. G. Hafez, Factor-bridging algorithm for the prediction of job satisfaction: Developing country perspective, J. King Saud Univ. Comput. Inf. Sci., 35 (2023), 101743. https://doi.org/10.1016/j.jksuci.2023.101743 doi: 10.1016/j.jksuci.2023.101743
[24]	R. Wu, X. Zhao, Z. Li, Y. Xie, The role of employee personality in employee satisfaction and turnover: Insights from online employee reviews, Pers. Rev., 53 (2024), 1581–1611. https://doi.org/10.1108/pr-04-2023-0309 doi: 10.1108/pr-04-2023-0309
[25]	S. Fraihat, Q. Shambour, M. A. Al-Betar, S. N. Makhadmeh, Variational autoencoders-based algorithm for multi-criteria recommendation systems, Algorithms, 17 (2024), 561. https://doi.org/10.3390/a17120561 doi: 10.3390/a17120561
[26]	N. Milano, M. Casella, R. Esposito, D. Marocco, Exploring the potential of variational autoencoders for modeling nonlinear relationships in psychological data, Behav. Sci., 14 (2024), 527. https://doi.org/10.3390/bs14070527 doi: 10.3390/bs14070527
[27]	H. Yang, R. Rang, L. Xing, L. Zhang, H. Cai, M. Guo, et al., V-GMR: A variational autoencoder-based heterogeneous graph multi-behavior recommendation model, Appl. Intell., 54 (2024), 3337–3350. https://doi.org/10.1007/s10489-024-05360-x doi: 10.1007/s10489-024-05360-x
[28]	M. Celia, S. Monaco, D. Apiletti, A comparative study of neural ordinary differential equations and neural operators for modeling temporal dynamics, Neural Comput. Applic., 37 (2025), 25319–25338. https://doi.org/10.1007/s00521-025-11580-0 doi: 10.1007/s00521-025-11580-0
[29]	Z. Chang, S. Liu, R. Qiu, S. Song, Z. Cai, G. Tu, Time-aware neural ordinary differential equations for incomplete time series modeling, J. Supercomput., 79 (2023), 18699–18727. https://doi.org/10.1007/s11227-023-05327-8 doi: 10.1007/s11227-023-05327-8
[30]	F. Zhan, X. Zhou, S. Li, D. Jia, H. Song, Learning latent ODEs with graph RNN for multi-channel time series forecasting, IEEE Signal Process. Lett., 30 (2023), 1432–1436. https://doi.org/10.1109/LSP.2023.3320439 doi: 10.1109/LSP.2023.3320439

Reader Comments

Your name:*

Email:*
© 2026 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)