Create weights based on various predictions, potentially using their errors with given constraints

Source: R/weight_functions.R

Create weights based on various predictions, potentially using their errors with given constraints

ensemble_weights(return_predictions, errors, method = "simple_average")

Arguments

return_predictions

tibble of stock_id, date, and various predictions

errors

tibble of stock_id, date, and various prediction errors

method

method to use for ensemble weights. Possible values are:

  • simple_average: simple average of predictions

  • weighted_average: weighted average based on the inverse of average errors

  • error_covariance: using the covariance of errors to calculate weights

Value

tibble with stock_id, date, and weight

Examples

if (FALSE) {
}