员工流动季节性问题及SARIMA-ANN模型应用

doi:10.12005/orms.2021.0053

摘要/Abstract

摘要： 中国企业员工流动受到中国商品生产季节性、劳动力市场供给周期性、员工身份两栖化等多种因素的影响,其数据呈现出季节性、非线性等特征。单一的自回归单整移动平均模型(ARIMA模型)不能较好地拟合其发展趋势并预测未来。本文通过神经网络模型(ANN模型)来修正传统的自回归单整移动平均模型(ARIMA模型),并加入季节性因素,从而形成SARIMA-ANN耦合模型,对企业员工流动的数据进行拟合与预测。通过对多组SARIMA-ANN模型的构建、衡量、比较与讨论,最终确定了较佳的神经网络对时间序列模型进行修正的耦合模型。实证结果显示,SARIMA-ANN模型充分考虑数据的季节性与趋势性,随机性与非线性特征并存的问题,对于季节性时间序列的经济数据的处理与预测是切实可行的。该耦合模型的应用证实了中国企业员工流动数据的趋势性与季节性、线性与非线性特征并存。这说明中国企业员工的流动具有更复杂的不规则的运动与突变,在精确预测有一定难度的情况下做好现有员工的留存工作是首要之策。

Abstract: The data and its extension tendency of employee turnover in Chinese enterprises cannot be well fitted and predicted by a single Autoregressive Integrated Moving-average Model (ARIMA Model) because of the seasonality and non-linearity of the data. In this paper, the SARIMA-ANN coupled model formed through the traditional ARIMA Model is modified by artificial neural network model (ANN). And seasonal factors and the data and its extension tendency of employee turnover are fitted and predicted by the SARIMA-ANN coupled model. A case study with the data of 93 months in a Chinese matured enterprise is used to validate the coupled model. The best coupling model is finally determined through the construction, measurement, comparison and discussion of multi-group SAARIMA-ANNA models. The results of the case study show that SARIMA-ANN model is feasible to fit and predict economic data of seasonal time series because the model fully deal with the coexistence of the seasonality and tendency, randomness and non-linearity of data. The application of the coupling model verifies the existence of this feature, that is, the data of employee turnover in Chinese enterprises coexist with the tendency and seasonality, linearity and non-linearity. This indicates that the employee turnover in Chinese enterprises has more complicated irregular movement and mutation, and it is the first step to do a good job in retaining existing employees when it is difficult to accurately predict.

Key words: employee turnover, the time series, SARIMA, artificial neural network

中图分类号:

F224.9

彭丽君, 马跃如. 员工流动季节性问题及SARIMA-ANN模型应用[J]. 运筹与管理, 2021, 30(2): 139-145.

PENG Li-jun, MA Yue-ru. SARIMA-ANN Model Application in Seasonal Problem of Employee Turnover[J]. Operations Research and Management Science, 2021, 30(2): 139-145.

参考文献

[1] AnnDenvir, Frank McMahon. Labour turnover in london hotels and the cost effectiveness of preventative measures[J]. International Journal of Hospitality Management, 1992, 11(2): 143-154.
[2] John R. Harris, Michael P. Todaro. Migration, unemployment and development: a two-sector analysis[J]. American Economic Review, 1970. 126-142.
[3] Griffeth R W, Hom P W. Retaining valued employees[M]. Sage Publications, 2001.
[4] Dustmann C, Weiss Y. Return migration: theory and empirical evidence from the UK[J]. British Journal of Industrial Relations, 2007, 45(2): 236-256.
[5] 姚先国,乔明睿,来君.城乡劳动力流动模式的影响因素分析——基于杭州市外来务工人员的调查[J].重庆大学学报:社会科学版,2009,15(1):14-19.
[6] 黄宇虹,弋代春,揭梦吟.中国小微企业员工流动现状、作用及其影响因素分析[J].管理世界,2016,12:77-89.
[7] March J G, Simon H A. Organizations[M]. Wiley, New York, 1958.
[8] Lee T W, Mitchell T R, Holtom B C, et al. Theoretical development and extension of the unfolding model of voluntary turnover[J]. Academy of Management Journal, 1999, 42(4): 450-462.
[9] Hylleberg S. General introduction//Hyllegerg S. Modeling seasonality[C]. Oxford: Oxford University Press, 1992.3-14.
[10] Ghiselli E E. Some perspective for industrial psychology[J]. American Psychologist, 1974, 29: 80-87.
[11] Shaw Jason D, Gupta Nina, Delery John E. Alternative conceptualizations of the relationship between voluntary turnover and organizational performance[J]. Academy of management journal, 2005, 48(1): 50-68.
[12] 肖良.基于季节性ARIMA模型的居民消费水平预测[J].统计与决策,2016,8:83-86.
[13] Mehdi Khashei, Mehdi Bijari. An artificial neural network(p,d,q)model for timeseries forecasting[J]. Expert Systems with applications, 2010, 37(1): 479-489.
[14] Mehdi Khashei, Mehdi Bijari. A novel hybridization of artificial neural networks and ARIMA models for time series forecasting[J]. Applied Soft Computing, 2011, 11(2): 2664-2675.
[15] Hassan Mohammadi, Daniel P. Rich. Dynamics of unemployment insurance claims: an application of ARIMA-GARCH models[J]. Atlantic Economic Journal, 2013, 41(4): 413-425.
[16] Kwangbok Jeong, Choongwan Koo, Taehoon Hong. An estimation model for determining the annual energy cost budget in educational facilities using SARIMA (seasonal autoregressive integrated moving average)and ANN(artificial neural network)[J]. Energy, 2014, 71: 71-79.
[17] 沈杰,杨忠月,乔吉良.基于季节性时间序列模型的我国林业产值预测分析[J].南京林业大学学报(自然科学版),2018,42.
[18] Peter Zhang G. Time series forecasting using a hybrid ARIMA and neural network model[J]. Neurocomputing, 2003, 50: 159-175.
[19] Fang-MeiTseng, Hsiao-ChengYu, Gwo-HsiungTzeng. Combining neural network model with seasonal time series ARIMA model[J]. Technological Forecasting and Social Change, 2002, 69(1): 71-87.
[20] 龙勇,苏振宇,汪於.基于季节调整和BP神经网络的月度负荷预测[J].系统工程理论与实践,2018,38(4):1052-1060.
[21] Ma Y, Peng L. Coupled model of artificial neural network and grey model for tendency prediction of labor turnover[J]. Mathematical Problems in Engineering, 2014, 2014.
[22] Ruiz-Aguilar J J, Turias I J, Jiménez-Come M J. Hybrid approaches based on SARIMA and artificial neural networks for inspection time series forecasting[J]. Transportation Research Part E: Logistics and Transportation Review, 2014, 67: 1-13.