FFORMA: Feature-based forecast model averaging

By Pablo Montero-Manso, George Athanasopoulos, Rob J Hyndman, T. Talagala. 2020

310 citations recorded by the source corpus.

We propose an automated method for obtaining weighted forecast combinations using time series features. The proposed approach involves two phases. First, we use a collection of time series to train a meta-model for assigning weights to various possible forecasting methods with the goal of minimizing the average forecasting loss obtained from a weighted forecast combination. The inputs to the meta-model are features that are extracted from each series. Then, in the second phase, we forecast new series using a weighted forecast combination, where the weights are obtained from our previously trained meta-model. Our method outperforms a simple forecast combination, as well as all of the most popular individual methods in the time series forecasting literature. The approach achieved second position in the M4 competition.

Subfields: Management Science and Operations Research, Signal Processing, Time-Series & Forecasting

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Management Science and Operations Research

In Signal Processing

In Time-Series & Forecasting

Citations in the corpus

N-BEATS: Neural basis expansion analysis for interpretable time series forecasting (2019, background): “In fact, 2 out of top-5 entries in the M4 competition are approaches of this type, including the second entry [24].”
Spectral Temporal Graph Neural Network for Multivariate Time-series Forecasting (2020, background): “Time-series forecasting is an emerging topic in machine learning, which can be divided into two major categories: univariate techniques [20, 22, 18, 27, 32, 19, 18] and multivariate techniques [24, 21, 17, 31, 3, 29, 25, 16, 15].”
Forecast combinations: An over 50-year review (2022, methodology): “…et al., 2017a; Petropoulos and Svetunkov, 2020) to sophisticated methods that tailor weights for different individual models (e.g., Bates and Granger, 1969; Newbold and Granger, 1974; Kolassa, 2011; Li et al., 2020; Montero-Manso et al., 2020; Kang et al., 2021; Wang et al., 2022b).”
Recurrent Neural Networks for Time Series Forecasting: Current Status and Future Directions (2019, methodology): “Although the state of the art in forecasting can now be seen as the methods of the M4 competition winners, most notably Smyl (2018) and Montero-Manso et al. (2018), we do not present comparisons against those methods as they are not automated in a way that they would be straightforwardly applicable…”
Forecasting in social settings: The state of the art (2020, methodology): “Progress has been made in this area too, with the recent FFORMA method (Montero-Manso et al. 2019) providing average coverage of 94.”