FFORMA: Feature-based Forecast Model Averaging

 

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 to assign 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 extracted from each series. 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, and outperforms all of the most popular individual methods in the time series forecasting literature. The approach achieved second position in the M4 competition.

Associated R package: fforma

Cite

@article{fforma,
  author  = {Pablo Montero-Manso and George Athanasopoulos and Rob J Hyndman and Thiyanga S Talagala},
  doi     = {10.1016/j.ijforecast.2019.02.011},
  journal = {International J Forecasting},
  number  = {1},
  pages   = {86-92},
  title   = {{FFORMA}: Feature-based Forecast Model Averaging},
  url     = {http://robjhyndman.com/publications/fforma/},
  volume  = {36},
  year    = 2020
}