George Athanasopoulosa, Rob J. Hyndmana, Nikolaos Kourentzesb, Fotios Petropoulosc

a Department of Econometrics and Business Statistics, Monash University, Australia
b Lancaster University Management School, Department of Management Science, UK
c School of Management, University of Bath, UK

European Journal of Operational Research (2017), to appear.

This paper introduces the concept of Temporal Hierarchies for time series forecasting. A temporal hierarchy can be constructed for any time series by means of non-overlapping temporal aggregation. Predictions constructed at all aggregation levels are combined with the proposed framework to result in temporally reconciled, accurate and robust forecasts. The implied combination mitigates modelling uncertainty, while the reconciled nature of the forecasts results in a unified prediction that supports aligned decisions at different planning horizons: from short-term operational up to long-term strategic planning. The proposed methodology is independent of forecasting models. It can embed high level managerial forecasts that incorporate complex and unstructured information with lower level statistical forecasts. Our results show that forecasting with temporal hierarchies increases accuracy over conventional forecasting, particularly under increased modelling uncertainty. We discuss organisational implications of the temporally reconciled forecasts using a case study of Accident & Emergency departments.

Keywords: Hierarchical forecasting, temporal aggregation, reconciliation, forecast combination

Download working paper

Online paper

  Tag: seasonality

1 posts
February 28th, 2017

Forecasting with temporal hierarchies

George Athanasopoulosa, Rob J. Hyndmana, Nikolaos Kourentzesb, Fotios Petropoulosc a Department of Econometrics and Business Statistics, Monash University, Australia b […]

January 13th, 2017

Visualising forecasting algorithm performance using time series instance spaces

Yanfei Kang1, Rob J Hyndman2, Kate Smith-Miles3 School of Statistics, Renmin University of China. Department of Econometrics and Business Statistics, […]

September 15th, 2016

Forecasting large collections of related time series

Talk given at the German Statistical Week, Augsburg, 15 September 2016

August 22nd, 2016

thief package for R

The thief package provides tools for Temporal Hierarchical Forecasting. The methods are described in Forecasting with temporal hierarchies, co-authored with […]

May 25th, 2016

ISCRR time series workshop

Materials for one-day workshop on time series and forecasting presented to ISCRR, Melbourne. Textbook Hyndman and Athanasopoulos (2014): Software […]

January 1st, 2016

Bagging exponential smoothing methods using STL decomposition and Box-Cox transformation

Christoph Bergmeir1, Rob J Hyndman2, José M Benítez1 Department of Computer Science and Artificial Intelligence, University of Granada, Spain. Department […]

June 23rd, 2015

MEFM: An R package for long-term probabilistic forecasting of electricity demand

International Symposium on Forecasting Riverside, California   I will describe and demonstrate a new open-source R package that implements the […]

June 19th, 2015

Probabilistic forecasting of peak electricity demand

Southern California Edison Rosemead, California   Electricity demand forecasting plays an important role in short-term load allocation and long-term planning […]

June 8th, 2015

STR: A Seasonal-Trend Decomposition Procedure Based on Regression

By Alex Dokumentov and Rob J Hyndman

June 1st, 2015

Large-scale unusual time series detection

Rob J Hyndman, Earo Wang and Nikolay Laptev

May 26th, 2015

Visualization of big time series data

Talk given to a joint meeting of the Statistical Society of Australia (Victorian branch) and the Melbourne Data Science Meetup Group.

February 23rd, 2015

Visualization and forecasting of big time series data

Talk given at the ACEMS Big data workshop, QUT.

September 23rd, 2014

Forecasting: principles and practice (UWA course)

Workshop held at UWA on 23-25 September 2014.

March 27th, 2013

bfast: Breaks For Additive Season and Trend

Automatic identification of breaks for additive season and trend, designed for use with remote sensing data. Based on Verbesselt, Hyndman, […]

December 31st, 2011

Forecasting time series with complex seasonal patterns using exponential smoothing

Alysha M De Livera, Rob J Hyndman and Ralph D Snyder Journal of the American Statistical Association (2011) 106(496), 1513-1527. […]

August 17th, 2010

Phenological change detection while accounting for abrupt and gradual trends in satellite image time series

Jan Verbesselt1, Rob J Hyndman2, Achim Zeilis3, Darius Culvenor1 Remote sensing team, CSIRO Sustainable Ecosystems, Private Bag 10, Melbourne VIC 3169, Australia Department of […]

January 15th, 2010

Detecting trend and seasonal changes in satellite image time series

Jan Verbesselt1, Rob J Hyndman2, Glenn Newnham1, Darius Culvenor1 Remote Sensing of Environment (2010), 114(1), 106-115. Remote sensing team, CSIRO […]

October 16th, 2004

The interaction between trend and seasonality

International Journal of Forecasting (2004), 20(4), 561-563. (A contribution to the discussion of Miller and Williams (2004).) Rob J. Hyndman […]

August 9th, 2000

Seasonal adjustment methods for the analysis of respiratory disease in environmental epidemiology

Bircan Erbas1 and Rob J Hyndman2 Department of General Practice & Public Health, The University of Melbourne, VIC 3010, Australia. […]