train_fforma_model: Train the FFORMA model

In ing-bank/tsforecast: Time Series Forecasting Pipeline

Description

train_fforma_model is a function to train a Feature-based Forecast Model Averaging (FFORMA) model, which weights the forecasts from the different available forecast models within the tsforecast package to create a new ensamble forecast. The FFORMA model uses the characteristics from the time series objects to predict which forecast model gives the best forecast performance. After training on all available training sets within the specified directory, the resulting FFORMA model is stored within the package to be reused at runtime during time series forecasting.

Usage

train_fforma_model(
  dir_fforma = file.path(getwd(), "data/fforma_training_data"),
  verbose = TRUE
)

Arguments

dir_fforma	A character string specifying the path to the directory where the training data os stored to do FFORMA training.
verbose	Boolean, which is set to TRUE if status updates are valued, or set to FALSE if they are not.

Value

A FFORMA meta learning model, trained according to the approach descibed in this paper and this example on github.

Examples

train_fforma_model()

ing-bank/tsforecast documentation built on Sept. 18, 2020, 9:40 a.m.