Method
- Divide the available data into training and validation sets.
- Use a large number of hidden units.
- Use very small random initial values.
- Use a slow learning rate.
- Compute the validation error rate periodically during training.
- Stop training when the validation error rate "starts to go up".
It is crucial to realize that the validation error is not a good estimate of the generalization error. One method for getting an unbiased estimate of the generalization error is to run the net on a third set of data, the test set, that is not used at all during the training process. The error on the test set gives estimate on generalization; to have the outputs of the net approximate target values given inputs that are not in the training set.
Read more about this topic: Early Stopping
Famous quotes containing the word method:
“Protestantism has the method of Jesus with His secret too much left out of mind; Catholicism has His secret with His method too much left out of mind; neither has His unerring balance, His intuition, His sweet reasonableness. But both have hold of a great truth, and get from it a great power.”
—Matthew Arnold (1822–1888)
“I have a new method of poetry. All you got to do is look over your notebooks ... or lay down on a couch, and think of anything that comes into your head, especially the miseries.... Then arrange in lines of two, three or four words each, don’t bother about sentences, in sections of two, three or four lines each.”
—Allen Ginsberg (b. 1926)
“The method of political science ... is the interpretation of life; its instrument is insight, a nice understanding of subtle, unformulated conditions.”
—Woodrow Wilson (1856–1924)