Hierarchical time series forecasting with an application to traffic accident counts

계층적 시계열 분석을 이용한 지역별 교통사고 발생건수 예측

  • Lee, Jooeun (Department of Applied Statistics, Chung-Ang University) ;
  • Seong, Byeongchan (Department of Applied Statistics, Chung-Ang University)
  • 이주은 (중앙대학교 응용통계학과) ;
  • 성병찬 (중앙대학교 응용통계학과)
  • Received : 2017.01.05
  • Accepted : 2017.01.23
  • Published : 2017.02.28


The paper introduces bottom-up and optimal combination methods that can analyze and forecast hierarchical time series. These methods allow forecasts at lower levels to be summed consistently to upper levels without any ad-hoc adjustment. They can also potentially improve forecast performance in comparison to independent forecasts. We forecast regional traffic accident counts as time series data in order to identify efficiency gains from hierarchical forecasting. We observe that bottom-up or optimal combination methods are superior to independent methods in terms of forecast accuracy.

본 논문에서는 계층적 시계열 자료 분석을 위한 대표적인 두 가지 방법인 상향식과 최적조합 예측법을 소개한다. 이러한 예측법은 계층적 시계열을 구성하는 모든 계열을 예측해야 하는 독립적 예측과 달리, 임의의 조정 과정이 없이 하위 계층 계열의 예측값의 합은 항상 상위 계층의 예측값과 일치하게 된다. 또한, 독립적 예측과 비교하여 예측력을 향상시킨다. 계층적 예측법의 효율성을 살펴보기 위하여 국내 16개 시도별 남녀 교통사고 발생건수 시계열 자료를 예측하였다. 이를 통하여 교통사고 발생건수에 대한 각 계층의 예측에서 계층적 방법과 독립적 방법의 차이점 및 우수성을 비교하였다.


Supported by : 중앙대학교


  1. Athanasopoulos, G., Ahmed, R. A., and Hyndman, R. J. (2009). Hierarchical forecasts for Australian domestic tourism, International Journal of Forecasting, 25, 146-166.
  2. Demsar, J. (2006). Statistical comparisons of classifiers over multiple data sets, Journal of Machine Learning Research, 7, 1-30.
  3. Friedman, M. (1937). The use of ranks to avoid the assumption of normality implicit in the analysis of variance, Journal of the American Statistical Association, 32, 675-701.
  4. Han, S. J. (2007). Road accident characteristics in metropolitan cities and provinces, Journal of Environmental Studies, 46, 211-220.
  5. Han, S. J. and Kim, K. J. (2007). Road accident trends analysis with time series models for various road types, International Journal of Highway Engineering, 9, 1-12.
  6. Hyndman, R. J. (2016). Forecast: Forecasting functions for time series and linear models, R package version 7.3, Available from:
  7. Hyndman, R. J., Ahmed, R. A., Athanasopoulos, G., and Shang, H. L. (2011). Optimal combination forecasts for hierarchical time series, Computational Statistics and Data Analysis, 55, 2579-2589.
  8. Hyndman, R. J. and Athanasopoulos, G. (2014). Forecasting Principles and Practice, Otexts, Available from:
  9. Hyndman, R. J, Wang, E., Lee, W., and Wichramasuriya, S. (2016). hts: Hierarchical and grouped time series, R package version 5.0, Available from:
  10. Kim, Y. S. and Lee, M. J. (2014). The analysis of predicting traffic accident using ARIMA model, Proceeding of the Korean Society of Civil Engineers Autumn Conference, 705-706.
  11. Nemenyi, P. B. (1963). Distribution-free Multiple Comparisons (PhD thesis), Princeton University, New Jersey.
  12. Pohlert, T. (2016). PMCMR: Calculate pairwise multiple comparisons of mean rank sums, R package version 4.1, Available from:
  13. Shang, H. L and Smith, P. W. F. (2013). Grouped time-series forecasting with an application to regional infant mortality counts, Southampton, GB, ESRC Centre for Population Change 40.
  14. Weale, M. (1988). The reconciliation of values, volumes and prices in the national accounts, Journal of the Royal Statistical Society, Series A, 151, 211-221.
  15. Zellner, A. and Tobias, J. (2000). A note on aggregation, disaggregation and forecasting performance, Journal of Forecasting, 19, 457-469.<457::AID-FOR761>3.0.CO;2-6