Evaluation of the best M4 competition methods for small area population forecasting

Wilson, Tom; Grossman, Irina; Temple, Jeromey

doi:10.1016/j.ijforecast.2021.09.005

Cited by 6 publications

(4 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Across all models, forecasts are generally more accurate and have better Coverage for regional areas than for major cities or more remote areas. Similar results were found in an evaluation of deterministic methods for Australian SA2s in Wilson et al, (2021b), however, results were different for New Zealand with Major Urban and Rural areas generally having smaller MedAPEs than Medium and Small urban areas. Whilst our overall results indicate that the addition of GBMs to ensemble models improves forecast performance, analysis by remoteness found that this is only consistently true for major cities.…”

Section: Discussionsupporting

confidence: 78%

“…The overall MedAPEs of the Ensemble models were considerably lower than those of the individual models. Indeed, the ALL ensemble produced more accurate point forecasts in terms of MedAPE than all bar one of the unconstrained results found in an evaluation of the top M4 competition methods for point forecasts of Australian SA2 small areas (Wilson et al, 2021b). Overall, the ALL average model performed better than the STAT model, indicating that it was advantageous to include the GBMs in the ensemble.…”

Section: Discussionmentioning

confidence: 87%

“…Previous research has shown that the accuracy of deterministic forecasts tends to be less for remote and very remote areas (Wilson and Rowe, 2011; Wilson et al, 2021b). We investigated whether remoteness influences the performance of both point forecasts and interval forecasts.…”

Section: Resultsmentioning

confidence: 99%

See 2 more Smart Citations

Development and evaluation of probabilistic forecasting methods for small area populations

Grossman

Bandara

Wilson

et al. 2023

Environment and Planning B: Urban Analytics and City Science

Self Cite

View full text Add to dashboard Cite

Planning and development decisions in both the government and business sectors often require small area population forecasts. Unfortunately, current methods often produce forecasts that are inaccurate, particularly for remote areas and those with smaller populations. Such inaccuracy necessitates the development and evaluation of methods to forecast and communicate forecast uncertainty, however, little research has been conducted in this domain for small area populations. In this paper, we evaluate a set of probabilistic forecasting methods which include Autoregressive integrated moving average, Exponential Smoothing, THETA, LightGBM and XGBOOST, to produce point forecasts and 80% prediction intervals for Australian SA2 small area populations. We also investigate methods to combine the intervals to produce ensemble forecasts. Our results show that individual probabilistic methods generally produce prediction intervals which underestimate forecast uncertainty. Combining forecasts improves the overall accuracy of point forecasts and the coverage of their intervals, however, coverage still tends to be less than the expected 80% for all but the most conservative combination method.

show abstract

Section: Discussionsupporting

confidence: 78%

Section: Discussionmentioning

confidence: 87%

See 1 more Smart Citation

Development and evaluation of probabilistic forecasting methods for small area populations

Grossman

Bandara

Wilson

et al. 2023

Environment and Planning B: Urban Analytics and City Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…It is also important to reduce the number of bad forecasts. To evaluate the number of bad forecasts by our models we define the Percentage of Bad Forecasts (Wilson et al 2021b) as the percentage of forecasts which have an Absolute Percentage Error greater than 10% after 5 years, or greater than 20% after 10 years. These values were selected because they are larger than the levels of error acceptable to population forecast users surveyed by Wilson and Shalley (2019).…”

Section: Error Metricsmentioning

confidence: 99%

Forecasting small area populations with long short-term memory networks

Grossman

Wilson

Temple

2023

Socio-Economic Planning Sciences

Self Cite

View full text Add to dashboard Cite

Population Fusion Transformer for Subnational Population Forecasting

Alghanmi,

Alotaibi,

Alshammari

et al. 2024

Int J Comput Intell Syst

View full text Add to dashboard Cite

Forecasting the subnational population accurately is needed for sustainable development, including planning for the future, allocating resources, or providing health services. Two approaches are used for forecasting subnational populations: local forecasting where a model is trained for each area, and global forecasting, where one model is trained with all areas. Local forecasting (e.g., statistical models) is limited to capturing the population growth patterns in a single area. Machine learning models, such as the light gradient boosting model (LGBM), are considered a more suitable approach for global forecasting, but it is limited to one-step predictions, leading to error accumulation. Also, combining several models into one ensemble model are used which helped in reduce forecasting errors. However, the nature of population growth is nonlinear, and there is a need to reduce error accumulation. This study overcomes these issues and proposes a population fusion transformer (PFT) as a global forecasting model for population forecasting, which outputs multi-step predictions. The PFT is based on a temporal fusion transformer (TFT) proposing a novel deep gated residual network (DGRN) block to capture data nonlinearity. This study also incorporates the proposed PFT model into various ensemble models to reduce forecasting errors using different prediction and learning approaches. The proposed models are applied to four subnational population datasets from several countries. The PFT model outperforms the LGBM and TFT with lower forecasting errors in three and two datasets. More importantly, combining the PFT with other models in ensemble models reduced errors further.

show abstract

Evaluation of the best M4 competition methods for small area population forecasting

Cited by 6 publications

References 42 publications

Development and evaluation of probabilistic forecasting methods for small area populations

Development and evaluation of probabilistic forecasting methods for small area populations

Forecasting small area populations with long short-term memory networks

Population Fusion Transformer for Subnational Population Forecasting

Contact Info

Product

Resources

About