Evaluate model performance
Now that you have both the actual and predicted values of each fold you can compare them to measure performance.
For this regression model, you will measure the Mean Absolute Error (MAE) between these two vectors. This value tells you the average difference between the actual and predicted values.
Cet exercice fait partie du cours
Machine Learning in the Tidyverse
Instructions
- Calculate the MAE by comparing the actual with the predicted values for the validate data and assign it to the
validate_maecolumn. - Print the
validate_maecolumn (note how they vary). - Calculate the mean of this column.
Exercice interactif pratique
Essayez cet exercice en complétant cet exemple de code.
library(Metrics)
# Calculate the mean absolute error for each validate fold
cv_eval_lm <- cv_prep_lm %>%
mutate(validate_mae = map2_dbl(___, ___, ~mae(actual = .x, predicted = .y)))
# Print the validate_mae column
cv_eval_lm$___
# Calculate the mean of validate_mae column
___