Tuning Parameters for Exponential Smoothing Models

Source: R/dials-exp_smoothing_params.R

Tuning Parameters for Exponential Smoothing Models

seasonality_type()

method()

error_method()

Value

A parameter

A parameter

A parameter

Details

The main parameters for Exponential Smoothing models are:

  • garch_order: Integer with the garch order.

  • arch_order: Integer with the arch_order.

  • mgarch_order: Integer with the mgarch order.

  • garch_t_student: A boolean value to specify for a generalized t-student garch model.

  • asymmetry: a string value for the asymmetric function for an asymmetric GARCH process. By default the value "none" for standard GARCH process. If "logit" a logistic function is used for asymmetry, and if "exp" an exponential function is used.

  • non_seasonal_ar: The order of the non-seasonal auto-regressive (AR) terms.

  • non_seasonal_ma: The order of the non-seasonal moving average (MA) terms.

  • markov_chains: The number of markov chains.

  • adapt_delta: The thin of the jumps in a HMC method

  • tree_depth: Maximum depth of the trees

Examples

non_seasonal_ar()

#> Non-seasonal AR Term (quantitative)
#> Range: [0, 5]

non_seasonal_differences()

#> Non-seasonal Differencing Term (quantitative)
#> Range: [0, 2]

non_seasonal_ma()

#> Non-seasonal MA Term (quantitative)
#> Range: [0, 5]