準確的需求預測可大幅提升公司之決策品質，然而對間歇性需求而言，由於觀察到的需求量包含高頻率的零值，建立適當的預測模型來解釋零膨脹現象極具挑戰性。本研究採用觀察驅動的零膨脹泊松（zero-inflated Poisson）和零膨脹負二項（zero-inflated negative binomial）自我迴歸模型分析與預測間歇性時間序列需求，並針對三種不同的建模方法（含零膨脹計數時間序列模型、自我迴歸綜合移動平均模型和 Croston 模型），使用分類樹機器學習模型來識別影響模型預測誤差的關鍵特徵，並得出選擇合適模型的重要指南。研究結果指出零膨脹泊松和零膨脹負二項自我迴歸模型在間歇性需求預測建模的有效性，本研究亦透過產業中的真實數據實證研究中所得之六個模型選擇規則可提高整體預測準確性。

Accurately predicting future demand is a valuable asset to a company. However, for
intermittent demand, where the observed numbers of demand contain a high frequency
of zeros, building an appropriate forecasting model to account for the zero-inflated
phenomenon can be very challenging. In this study, we investigate and forecast
intermittent time series demand using observation-driven zero-inflated Poisson (ZIP) and zero-inflated negative binomial (ZINB) autoregressive models. We also employ
classification tree learning to identify key features that affect the magnitude of prediction errors (for different modeling approaches—zero-inflated count time series model, autoregressive integrated moving average (ARIMA) model, and Croston model) to
derive important guidelines for choosing the most appropriate approach for building a
forecasting model for intermittent demand data. We demonstrated the effectiveness of
ZIP and ZINB autoregressive models in dealing with intermittent demand forecasting.
Six rules for model selection have also been shown to improve the overall prediction
accuracy over real-world multi-item products.

Aljawarneh, S., Yassein, M.B. & Aljundi, M. (2019). An enhanced J48 classification algorithm for the anomaly intrusion detection systems. Cluster Computing. 22(5). 10549–10565. https://doi.org/10.1007/s10586-017-1109-8.
Box, G. & Jenkins, G. (1970). Time series analysis forecasting and control. Journal of Time Series Analysis. 3(3228). https://dx.doi.org/10.2307/1912100.Boylan, J.E., & Syntetos. A.A. (2007). The accuracy of a modified Croston procedure. International Journal of Production Economics. 107(2). 511-517. https://doi.org/10.1016/j.ijpe.2006.10.005.
Breiman, L., Friedman, J., Stone, C.J., & Olshen, R.A. (1984). Classification and regression trees. Andover, England, UK: Taylor & Francis. ISBN: 978-0-41204841-8.
Butler, K. (2019). Time series correlation and slopes with time [Course notes]. Retrieved from https://www.utsc.utoronto.ca/~butler/d29/time-series.pdf.
Christ, M., Braun, N., Neuffer, J., Kempa-Liehr, A.W. (2018). Time series FeatuRe Extraction on basis of scalable hypothesis tests (tsfresh – A Python package). Neurocomputing. 307(13). 72-77. https://doi.org/10.1016/j.neucom.2018.03.067.
Croston, J.D. (1972). Forecasting and stock control for intermittent demands. Journal of the Operational Research Society. 23. 289-303. https://doi.org/10.1057/jors.1972.50.
da Costa-Luis, C.O. (2019). tqdm: A fast, extensible progress meter for Python and CLI. The Journal of Open Source Software. 4(37). 1227. https://doi.org/10.21105/joss.01277.
Delgado-Bonal, A. & Marhsak, A. (2019). Approximate entropy and sample entropy: A comprehensive tutorial. Entropy. 21(6). 541. https://doi.org/10.3390/e21060541.
Descartes, R. (1954). The geometry of René Descartes: with a facsimile of the first edition (Smith, D.E. & Latham, M.L. (Trans.). Dover Publications, Inc. (Original work published 1637).
Fokianos, K. & Kedem, B. (1998). Prediction and classification of non-stationary categorical time series. Journal of Multivariate Analysis. 67. 277-296.
Harris, C.R., Millman, K.J., van der Walt, S.J., et al. (2020) Array programming with NumPy. Nature. 585. 357-362. https://doi.org/10.1038/s41586-020-2649-2.
Ho, S.L. & Xie, M. (1998). The use of ARIMA models for reliability forecasting and analysis. Computers & Industrial Engineering. 35. 213-216.
Hyndman, R.J. & Koehler, A.B. (2006). Another look at measures of forecast accuracy. International Journal of Forecasting. 22(4). 679-688. https://doi.org/10.1016/j.ijforecast.2006.03.001.Johnston, F.R., & Boylan, J.E. (1996). Forecasting for items with intermittent demand. Journal of the Operational Research Society. 47(1). 113-121.
Nikolopoulos, K., Syntetos A.A., Boylan, J.E., Petropoulos, F., & Assimakopoulos, V. (2011). An Aggregate-Disaggregate Intermittent Demand Approach (ADIDA) to forecasting: An empirical proposition and analysis. Journal of the Operational Research Society. 62(3). 544-554. https://doi.org/10.1057/jors.2010.32.
Kang, Y., Hyndman R.J., Smith-Miles, K. (2017). Visualising forecasting algorithm performance using time series instance spaces. International Journal of Forecasting. 33(2). 345-358. https://doi.org/10.1016/j.ijforecast.2016.09.004.
Kaya, G.O., Sahin, M. & Demirel, O.F. (2020). Intermittent demand forecasting: a guideline for method selection. Sādhanā. 45(51). https://doi.org/10.1007/s12046-020-1285-8.Kiefer, D., Grimm, F., Bauer, M., & Dinther, C. (2021). Demand forecasting intermittent and lumpy time series: Comparing statistical, machine learning and deep learning methods. 54th Hawaii International Conference on System Sciences. 1425-1434. https://dx.doi.org/ 10.24251/HICSS.2021.172.
Kourentzes, N. (2013). Intermittent demand forecasts with neural networks. International Journal of Production Economics. 143(1). 198-206. https://doi.org/10.1016/j.ijpe.2013.01.009.
Kourentzes, N. (2014). On intermittent demand model optimization and selection. International Journal of Production Economics. 156. 180-190. https://doi.org/10.1016/j.ijpe.2014.06.007.
Levén, E & Segerstedt, A. (2004). Inventory control with a modified Croston procedure and Erlang distribution. International Journal of Production Economics. 90(3). 361-367. https://doi.org/10.1016/S0925-5273(03)00053-7.
Lewis, R.J. (2000). An introduction to Classification and Regression Tree (CART) analysis. Annual Meeting of the Society for Academic Emergency Medicine. San Francisco, California.
Li, L., Kang, Y., Petropoulos, F., & Li, F. (2022). Feature-based intermittent demand forecast combinations: accuracy and inventory implications. International Journal of Production Research. https://doi.org/10.1080/00207543.2022.2153941.
Lubba, C.H., Sethi, S.S., Knaute, P. et al. (2019). catch22: CAnonical Time-series CHaracteristics. Data Mining Knowledge Discovery. 33, 1821–1852. https://doi.org/10.1007/s10618-019-00647-x.
Pearson, K. (1895). Contribution to the mathematical theory of evolution II. Skew variation in homogeneous material. Philosophical Transaction. 186. 1343-1414.
Petropoulos, F., Kourentzes, N., & Nikolopoulos, K. (2016). Another look at estimators for intermittent demand. International Journal of Production Economics. 181(A). 154-161. https://doi.org/10.1016/j.ijpe.2016.04.017.
Petropoulos, F., Makridakis, S. & Assimakopoulos, V. (2014). ‘Hourses for Courses’ in demand forecasting. European Journal of Operational Research. 237(1). 152-163. https://doi.org/10.1016/j.ejor.2014.02.036.Pincus, S. (1991). Approximate entropy as a measure of system complexity. Proceedings of the National Academy of Sciences of the United States of America. 88. 2297-2301. https://doi.org/10.1073/pnas.88.6.2297.
Python Data Analysis Library – pandas [online]. Retrieved from: https://pandas.pydata.org/.
Rao, A.V. (1973). A comment on: Forecasting and stock control for intermittent demands. Journal of the Operational Research Society. 24(4). 639-640. https://doi.org/10.1057/jors.1973.120.
Rožanec, J.M., Fortuna, B., & Mladenić, D. (2022). Reframing demand forecasting: A two-fold approach for lumpy and intermittent demand. Sustainability 2022. 14(15). 9295. https://doi.org/10.3390/su14159295.
Schmittlein, D. C., Bemmaor, A. C., and Morrison, D. G. (1985). Technical note—why does the nbd model work? Robustness in representing product purchases, brand purchases and imperfectly recorded purchases. Marketing Science. 4(3). 255-266.
Sci-kit learn: machine learning in Python [online]. Retrieved from: https://scikit-learn.org/stable/.Shenstone, L., & Hyndman, R.J. (2005). Stochastic models underlying Croston’s method for intermittent demand forecasting. Journal of Forecasting. 24(6). 389-402. https://doi.org/10.1002/for.963.
Smith, T.G., et al. (2017). pmdarima: ARIMA estimators for Python. http://www.alkaline-ml.com/pmdarima.
Syntetos, A. A., Babai, M. Z., & Altay, N. (2012). On the demand distributions of spare parts. International Journal of Production Research. 50(8). 2101–2117. https://doi.org/10.1080/00207543.2011.562561.
Syntetos, A.A. & Botlan, J.E. (2001). On the bias of intermittent demand estimates. International Journal of Production Economics. 71. 457-466. http://dx.doi.org/10.1016/S0925-5273(00)00143-2.
Syntetos, A.A. & Boylan, J.E. (2005). The accuracy of intermittent demand estimates. International Journal of Forecasting. 21(2). 303-314. https://doi.org/10.1016/j.ijforecast.2004.10.001.
Syntetos, M., Boylan, J., & Croston, J.D. (2005). On the categorization of demand patterns. Journal of the Operational Research Society. 56. https://doi.org/10.1057/palgrave.jors.2601841.
Tawiah, K., Iddrisu, W.A., & Asosega, K.A. (2021). Zero-inflated time series modelling of COVID-19 deaths in Ghana. Journal of Environmental and Public Health. https://doi.org/10.1155/2021/5543977.
Theodorou, E., Wang, S., Kang, Y., Spiliotis, E., Makridakis, S., & Assimakopoulos, V. (2022). Exploring the representativeness of the M5 competition data. International Journal of Forecasting. 38. 1500-1506. https://doi.org/10.1016/j.ijforecast.2021.07.006.
Tomcala, J. (2018). Time series entropies. https://cran.r-project.org/web/packages/TSEntropies/index.html.
Turrini, L. & Meissner, J. (2019). Spare parts inventory management: New evidence from distribution fitting. European Journal of Operational Research. 273(1). 118–130. https://doi.org/10.1016/j.ejor.2017.09.039.
Van Rossum, G. (2020). The Python library reference, release 3.8.2. Python Software Foundation.
Wickham, H., François, R., Henry, L., & Müller, K. (2022). dplyr: A grammar of data manupulation. https://dplyr.tidyverse.org, https://github.com/tidyverse/dplyr.
Willemain, T.R., Smart, C.N., Shockor, J.H., & DeSautels, P.A. (1994). Forecasting intermittent demand in manufacturing: a comparative evaluation of Croston’s method. International Journal of Forecasting. 10(4). 529-538. https://doi.org/10.1016/0169-2070(94)90021-3.
Williams, T.M. (1984). Stock control with sporadic and slow-moving demand. The Journal of the Operational Research Society. 35(10). 939-948. https://doi.org/10.2307/2582137.
Yang, M., Zamba, G.K.D., & Cavanaugh, J.E. (2013). Markov regression models for count time series with excess zeros: A partial likelihood approach. Statsitical Methodology. 14. 26-28. https://doi.org/10.1016/j.stamet.2013.02.001.
Yang, Y. (2005). Can the strengths of AIC and BIC be shared? A conflict between model indentification and regression estimation. Biometrika. 92(4). 937-950.
Zeileis, A., Kleiber, C., & Jackman, S. (2008). Regression model for count data in R. Journal of Statistical Software. 27(8). http://www.jstatsoft.org/v27/i08/.