Fast computation of reconciled forecasts for hierarchical and grouped time series

Hyndman, Rob J.; Lee, Alan J.; Wang, Earo

doi:10.1016/j.csda.2015.11.007

Cited by 132 publications

(100 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The corresponding GLS may be too large to solve without special numerical methods. Hyndman et al (2011, section 5) discuss some potential remedies, and Hyndman et al (2014a) have recently proposed a fast recursive approach for specific hierarchy topologies.…”

Section: Reconciliation Of Forecastsmentioning

confidence: 99%

Supply chain forecasting: Theory, practice, their gap and the future

Syntetos

Babai

Boylan

et al. 2016

European Journal of Operational Research

234

View full text Add to dashboard Cite

HIGHLIGHTSThe literature on supply chain forecasting is critically reviewed;The process of involving the forecasting community towards that task is described;Gaps between theory and practice are identified; Data and software related issues are explicitly considered; Challenges are summarised followed by suggestions for further research. ABSTRACTSupply Chain Forecasting (SCF) goes beyond the operational task of extrapolating demand requirements at one echelon. It involves complex issues such as supply chain coordination and sharing of information between multiple stakeholders. Academic research in SCF has tended to neglect some issues that are important in practice. In areas of practical relevance, sound theoretical developments have rarely been translated into operational solutions or integrated in state-of-the-art decision support systems. Furthermore, many experience-driven heuristics are increasingly used in everyday business practices. These heuristics are not supported by substantive scientific evidence; however, they are sometimes very hard to outperform. This can be attributed to the robustness of these simple and practical solutions such as aggregation approaches for example (across time, customers and products).This paper provides a comprehensive review of the literature and aims at bridging the gap between the theory and practice in the existing knowledge base in SCF. We highlight the most promising approaches and suggest their integration in forecasting support systems. We discuss the current challenges both from a research and practitioner perspective and provide a research and application agenda for further work in this area. Finally, we make a contribution in the methodology underlying the preparation of review articles by means of involving the forecasting community in the process of deciding both the content and structure of this paper.

show abstract

Section: Reconciliation Of Forecastsmentioning

confidence: 99%

Supply chain forecasting: Theory, practice, their gap and the future

Syntetos

Babai

Boylan

et al. 2016

European Journal of Operational Research

234

View full text Add to dashboard Cite

show abstract

“…The new combination approach proposed in Athanasopoulos et al (2009) andHyndman et al (2011) considers smaller hierarchies with less than four levels, while hierarchies with millions of time series at the lowest level are discussed in Hyndman et al (2014). Starting point of this approach is to forecast all time series at all (aggregate and disaggregate) levels first.…”

Section: Optimal Forecast Combination For Hierarchical Time Seriesmentioning

confidence: 99%

Combining country-specific forecasts when forecasting Euro area macroeconomic aggregates

Zeng

2016

Empirica

View full text Add to dashboard Cite

European Monetary Union (EMU) member countries' forecasts are often combined to obtain the forecasts of the Euro area macroeconomic aggregate variables. The aggregation weights which are used to produce the aggregates are often considered as combination weights. This paper investigates whether using different combination weights instead of the usual aggregation weights can help to provide more accurate forecasts. In this context, we examine the performance of equal weights, the least squares estimators of the weights, the combination method recently proposed by Hyndman et al. (2011) and the weights suggested by shrinkage methods. We find that some variables like real GDP and GDP deflator can be forecasted more precisely by using flexible combination weights. Furthermore, combining only forecasts of the three largest European countries helps to improve the forecasting performance. The persistence of the individual data seems to play an important role for the relative performance of the combination.

show abstract

“…On the other hand, it is possible that assumption (1) becomes less and less adequate, in particular for a longer and longer forecast 9 horizon. Hyndman et al (2016) proposed a GLS estimator, where the elements of Σ + h are set to the inverse of the variances of the base forecasts, Var(y n+1 − y n+1|n ). Note that we use the one-step-ahead forecast variances, not the h-step-ahead forecast variances.…”

Section: Optimal Combinationmentioning

confidence: 99%

“…This method is referred to as the middle-out method. Hyndman et al (2011) andHyndman, Lee andWang (2016) proposed an optimal combination method, where base forecasts are obtained independently for all series at all levels of the hierarchy and then a linear regression model is used with an ordinary least squares (OLS) or a generalized least squares (GLS) estimator to optimally combine and reconcile these forecasts. They showed that the revised forecasts do not only add up across the hierarchy, but they are also unbiased and have minimum variance amongst all combined forecasts under some simple assumptions (Hyndman et al, 2011).…”

Section: Introductionmentioning

confidence: 99%

Reconciling Forecasts of Infant Mortality Rates at National and Sub-National Levels: Grouped Time-Series Methods

Shang

2016

Popul Res Policy Rev

View full text Add to dashboard Cite

Mortality rates are often disaggregated by different attributes, such as sex, state, education, religion or ethnicity. Forecasting mortality rates at the national and subnational levels plays an important role in making social policies associated with the national and sub-national levels. However, base forecasts at the sub-national levels may not add up to the forecasts at the national level. To address this issue, we consider the problem of reconciling mortality rate forecasts from the viewpoint of grouped time-series forecasting methods (Hyndman, Ahmed, Athanasopoulos and Shang, 2011).A bottom-up method and an optimal combination method are applied to produce point forecasts of infant mortality rates that are aggregated appropriately across the different levels of a hierarchy. We extend these two methods by considering the reconciliation of interval forecasts through a bootstrap procedure. Using the regional infant mortality rates in Australia, we investigate the one-step-ahead to 20-step-ahead point and interval forecast accuracies among the independent and these two grouped time-series forecasting methods. The proposed methods are shown to be useful for reconciling point and interval forecasts of demographic rates at the national and sub-national levels, and would be beneficial for government policy decisions regarding the allocations of current and future resources at both the national and sub-national levels.

show abstract

Fast computation of reconciled forecasts for hierarchical and grouped time series

Cited by 132 publications

References 10 publications

Supply chain forecasting: Theory, practice, their gap and the future

Supply chain forecasting: Theory, practice, their gap and the future

Combining country-specific forecasts when forecasting Euro area macroeconomic aggregates

Reconciling Forecasts of Infant Mortality Rates at National and Sub-National Levels: Grouped Time-Series Methods

Contact Info

Product

Resources

About