Issues
- It's not clear on how many cases to assign to the training and validation sets
- The result might highly depend on the algorithm which is used to split the data into training and validation set
- Notion of "increasing validation error" is ambiguous; it may go up and down numerous times during training. The safest approach is to train to convergence, then determine which iteration had the lowest validation error. This impairs fast training, one of the advantages of early stopping.
Read more about this topic: Early Stopping
Famous quotes containing the word issues:
“Your toddler will be “good” if he feels like doing what you happen to want him to do and does not happen to feel like doing anything you would dislike. With a little cleverness you can organize life as a whole, and issues in particular, so that you both want the same thing most of the time.”
—Penelope Leach (20th century)
“I can never bring you to realize the importance of sleeves, the suggestiveness of thumb-nails, or the great issues that may hang from a boot-lace.”
—Sir Arthur Conan Doyle (1859–1930)
“The “universal moments” of child rearing are in fact nothing less than a confrontation with the most basic problems of living in society: a facing through one’s children of all the conflicts inherent in human relationships, a clarification of issues that were unresolved in one’s own growing up. The experience of child rearing not only can strengthen one as an individual but also presents the opportunity to shape human relationships of the future.”
—Elaine Heffner (20th century)