Pages
Techniques
Regularization
Regularization is the punishment of changes set to loss value to avoid params (weights and biases) change to much, leave space of possible values of params for unknown data, better for generalizing.
Regularisation technique is used to limit the weights and avoid them saturate into a large range and no possible to fit (match) with the inputs. It is done by using a portion of loss to punish (reduce) the changes to weights.
Want to print your doc?
This is not the way.
This is not the way.
Try clicking the ⋯ next to your doc name or using a keyboard shortcut (
CtrlP
) instead.