|
On Early Stopping in Gradient Descent Learning Abstractwe provide an error analysis to the gradient descent algorithm for learning and discuss connections of such a learning scheme with well known algorithms such as L2-boosting or landweber regularization
[Edit] |