Author(s)

HuiJun Liang^1#, ZiTian Sheng^2#, ZhiJie Luo^2*

Affiliation(s)

¹Guangzhou Port Shipping Co., Ltd, Guangzhou 510700, Guangdong, China.

²School of Artificial Intelligence, Zhongkai University of Agriculture and Engineering, Guangzhou 510225, Guangdong, China.

Corresponding Author

ZhiJie Luo

ABSTRACT

In the context of refined enterprise management, the dynamic optimal allocation of human resources has become the key to improving operational efficiency and reducing labor costs. Financial reimbursement, as a high-frequency and important business process, exhibits significant fluctuation in its workload due to influences from business cycles, project milestones, and other factors. A fixed number of finance personnel can easily lead to backlogs in expense review during peak periods, resulting in employee dissatisfaction, while creating idle human resources during off-peak periods. To resolve this contradiction, this paper aims to establish a dynamic adjustment mechanism for finance personnel based on predictive modeling. This study utilizes historical reimbursement claim volume time series data from enterprises, integrating the Prophet-LSTM predictive model based on the Holt-Winters algorithm to accurately predict future claim volumes for specific cycles (e.g., monthly, quarterly). Through this model, enterprise managers can proactively anticipate future workloads and dynamically adjust staffing levels for financial reimbursement positions accordingly, enabling flexible allocation of financial personnel. Empirical research demonstrates that this predictive model exhibits high accuracy, with an MSE of only 152.47, an RMSE as low as 12.35, an MAE of 7.89, a MAPE of just 6.58%, and an R2 as high as 0.948. The dynamic adjustment strategy based on this model effectively smooths workloads, significantly shortens reimbursement processing cycles, and enhances the overall utilization efficiency of corporate human resources. This study provides a feasible theoretical framework and practical pathway for enterprises to achieve data-driven human resource management in finance and other volatile business domains. 

KEYWORDS

Human resources; Financial reimbursement; Holt-Winters; Prophet-LSTM; Predictive model

CITE THIS PAPER

HuiJun Liang, ZiTian Sheng, ZhiJie Luo. Dynamic deployment of financial personnel based on the holt-winters algorithm-prophet-lstm predictive model. World Journal of Economics and Business Research. 2026, 4(1): 25-33. DOI: https://doi.org/10.61784/wjebr3085.

REFERENCES

[1] Ding W J, Zhang W. Research on the Paths of Enterprise Cost Reduction and Efficiency Improvement in the Digital Economy Era. Modernization of Management, 2020, 40(2): 88–91.

[2] Li J. Research on Operational Efficiency Optimization of Financial Shared Service Centers: From the Perspective of Process Reengineering. Friends of Accounting, 2020(12): 35–40.

[3] Smith J, Johnson L. Business Forecasting: A Practical Approach. Routledge, 2019.

[4] Atkinson J. Manpower strategies for flexible organisations. Personnel Management, 1984, 16(8): 28–31.

[5] Chen J, Wang T. Research on Flexible Allocation of Financial Human Resources Under the Shared Service Model. Human Resource Development of China, 2021, 38(5): 68–81.

[6] Wright P M, Snell S A. Toward a unifying framework for exploring fit and flexibility in strategic human resource management. Academy of Management Review, 1998, 23(4): 756–772.

[7] Davenport T H, Harris J G, Morison R. Analytics at Work: Smarter Decisions, Better Results. Harvard Business Press, 2010.

[8] Box G E P, Jenkins G M, Reinsel G C, et al. Time Series Analysis: Forecasting and Control (5th ed.). John Wiley & Sons, 2025.

[9] Hewamalage H, Bergmeir C, Bandara K. Global models for time series forecasting: A simulation study. Pattern Recognition, 2023, 134: 109121.

[10] Chatfield C. The Holt-Winters forecasting procedure. Journal of the Royal Statistical Society: Series C (Applied Statistics), 1978, 27(3): 264–279.

[11] Taylor S J, Letham B. Forecasting at scale. The American Statistician, 2018, 72(1): 37–45.

[12] Yu Y, Si X, Hu C, et al. A review of recurrent neural networks: LSTM cells and network architectures. Neural Computation, 2019, 31(7): 1235–1270.

[13] Sherstinsky A. Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network. Physica D: Nonlinear Phenomena, 2020, 404: 132306.

[14] García S, Ramirez-Gallego S, Luengo J, et al. Big data preprocessing: methods and prospects. Big Data Analytics, 2016, 1(1): 9.

[15] Deepa B, Ramesh K. Epileptic seizure detection using deep learning through min max scaler normalization. International Journal of Health Sciences, 2020, 1: 10981–10996.

[16] Bates J M, Granger C W J. The combination of forecasts. Journal of the Operational Research Society, 1969, 20(4): 451–468.

[17] Chicco D, Warrens M J, Jurman G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Computer Science, 2021, 7: e623.

[18] Nakagawa S, Schielzeth H. A general and simple method for obtaining R2 from generalized linear mixed-effects models. Methods in Ecology and Evolution, 2013, 4(2): 133–142.

[19] Press W H. Numerical Recipes 3rd Edition: The Art of Scientific Computing. Cambridge University Press, 2007.

[20] Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems, 2012, 25.