Forecast with forecasts: diversity matters
From MaRDI portal
Abstract: Forecast combinations have been widely applied in the last few decades to improve forecasting. Estimating optimal weights that can outperform simple averages is not always an easy task. In recent years, the idea of using time series features for forecast combination has flourished. Although this idea has been proved to be beneficial in several forecasting competitions, it may not be practical in many situations. For example, the task of selecting appropriate features to build forecasting models is often challenging. Even if there was an acceptable way to define the features, existing features are estimated based on the historical patterns, which are likely to change in the future. Other times, the estimation of the features is infeasible due to limited historical data. In this work, we suggest a change of focus from the historical data to the produced forecasts to extract features. We use out-of-sample forecasts to obtain weights for forecast combinations by amplifying the diversity of the pool of methods being combined. A rich set of time series is used to evaluate the performance of the proposed method. Experimental results show that our diversity-based forecast combination framework not only simplifies the modelling process but also achieves superior forecasting performance in terms of both point forecasts and prediction intervals. The value of our proposition lies on its simplicity, transparency, and computational efficiency, elements that are important from both an optimisation and a decision analysis perspective.
Recommendations
Cites work
- `Horses for courses' in demand forecasting
- A combination selection algorithm on forecasting
- An improved forecasting approach to reduce inventory levels in decentralized supply chains
- Combining ranked mean value forecasts
- Ensemble approaches for regression: a survey
- Forecaster Diversity and the Benefits of Combining Forecasts
- Forecasting time series with complex seasonal patterns using exponential smoothing
- Generalizing the Theta method for automatic forecasting
- GRATIS: GeneRAting TIme Series with diverse and controllable characteristics
- Measures of diversity in classifier ensembles and their relationship with the ensemble accuracy
- Review of guidelines for the use of combined forecasts
- Strictly Proper Scoring Rules, Prediction, and Estimation
- Supply chain forecasting when information is not shared
Cited in
(5)- Arbitrage of forecasting experts
- On the price of diversity for multiwinner elections under (weakly) separable scoring rules
- On the integration of multiple criteria decision aiding and forecasting: does it create value in portfolio selection?
- Meta-learning based on recurrent neural networks for ensembling forecasts of time series with multiple seasonal patterns
- An improved feature-based forecast combination method using rolling origin evaluation
This page was built for publication: Forecast with forecasts: diversity matters
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2140152)
