本研究探討計數型廣義線性模型於車流量之預測成效性。我們考慮 Poisson 時間序列模型以及 Conway-Maxwell-Poisson 迴歸模型，建立具有日內時段週期及週內週期兩種循環的週期性基底函數，並比較兩模型的預測結果。資料使用 2016 年 8 月至 2017 年 12 月高雄市中正一路與輔仁路口之車流量，利用預測日的前三週做為訓練樣本，預測未來 8 天的車流量，並以平均絕對比例誤差(Mean Absolute Percentage Errors, MAPE)當作評估模型的依據，最後結果可提供給相關單位未來車流量管制之參考。

This study discusses the applicability of the generalized linear model for count data in predicting the
flow in Kaohsiung area. We consider poisson time series model and Conway-Maxwell-Poisson (CMP) regression model proposed by Conway and Maxwell (1962) with periodic bases concerning the intra-daily and inter-weekly effects. Results obtained by the two methods are compared with traffic data on the intersection of the Zhongzhen First Road and Furen Road from Kaohsiung City traffic control, between August 2016 to December 2016, evaluated by the mean absolute percentage error (MAPE). More explicitly, the accuracy of the forecasting results for the forecasting day and the next seven days afterwards, based on traffic flow data three weeks right before the forecasting, will be presented throught the MAPE. The MAPE results show that the CMP model seem to have performed slightly better in general.

論文審定書 i
誌謝 ii
摘要 iii
Abstract iv
1 前言 1
2 資料描述 2
2.1 資料蒐集方式 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2.2 資料處理 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2.1 NA 值處理 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2.2 颱風天、特殊日與事故處理 . . . . . . . . . . . . . . . . . . . . . . . 4
2.2.3 時間軸合併 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 研究方法 6
3.1 Poisson 時間序列模型 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2 Conway-Maxwell-Poisson (CMP) 迴歸模型 . . . . . . . . . . . . . . . . . . 7
4 實證研究 9
4.1 Poisson 時間序列模型分析結果 . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.2 CMP 迴歸模型分析結果 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5 討論與結語 14
參考文獻 15
附錄 16
A 附錄 16
A.1 週期性函數 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
A.2 Conway-Maxwell-Poisson distribution . . . . . . . . . . . . . . . . . . . . . 18
A.2.1 Bernoulli 分佈 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
A.2.2 Poisson 分佈 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
A.2.3 Geomtric 分佈 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

[1] Fan, S., & Hyndman, R. J. (2012). Short-term load forecasting based on a semiparametric
additive model. IEEE Transactions on Power Systems, 27, 134-141.
[2] Fokianos, K., & Tjostheim, D. (2011). Log-linear Poisson autoregression. Journal of
Multivariate Analysis, 102, 563-578.
[3] Jung, R. C., & Tremayne, A. R. (2011). Useful models for time series of counts or
simply wrong ones?. AStA Advances in Statistical Analysis, 95, 59-91.
[4] King, A. A., Nguyen, D., & Ionides, E. L. (2016). Statistical inference for partially
observed Markov processes via the R package pomp. Journal of Statistical Software,
69, 1-43.
[5] Kumar, K., Parida, M., & Katiyar, V. K. (2013). Short term traffic flow prediction for
a non urban highway using artificial neural network. Procedia-Social and Behavioral
Sciences, 104, 755-764.
[6] Lewis, C. D. (1982). Industrial and business forecasting methods: A practical guide to
exponential smoothing and curve fitting. London: Butterworth.
[7] Liboschik, T., Fokianos, K., & Fried, R. (2017). tscount: An R package for analysis of
count time series following generalized linear models. Journal of Statistical Software,
82, 1-51.
[8] Piegl, L., & Tiller, W. (2012). The NURBS book. Springer: Berlin.
[9] Rodr´ıguez, G. (2013). Models for count data with overdispersion.
[10] Sellers, K. F., & Raim, A. (2016). A flexible zero-inflated model to address data
dispersion. Computational Statistics and Data Analysis, 99, 68-80.
[11] Sellers, K. F., & Shmueli, G. (2010). A flexible regression model for count data. The
Annals of Applied Statistics, 4, 943-961.
[12] Taylor, J. W. (2003). Short-term electricity demand forecasting using double seasonal
exponential smoothing. Journal of the Operational Research Society, 54, 799-805.