model_selection: Model Selection Constructor

In chrishaarstick/mad4sight: MadStork Foresight

Description

Helper function to create final model selection criteria for forecasting

Usage

model_selection(measure = "RMSE", n = 1, weights = "equal")

Arguments

measure	accuracy metric to arrange and weight by. Choices are RMSE and MAE. Default is RMSE
n	top_n number of models to select. Default is 1
weights	option to weight model forcasts by inverse of measure or equally. Options are equal for equal weighting or weighted for inverse weighting. Default is equal

Value

model selection object

Examples

# default
model_selection()

# weighted selection
model_selection(measure = "RMSE", n = 2, weights = "weighted")

chrishaarstick/mad4sight documentation built on May 6, 2019, 4:31 p.m.