Overfitting

What is Overfitting

Overfitting is when the neuralnet changes to adapt so much to a specific set of data.

Overfitting happens when

The number of training samples are too few

The number of neurons and layers are too much

Or both

Split training data into training set and test set.

To limit weight range into a small range, the rest of the range for weight values should be for those training sample unknown yet.

⁠

Want to print your doc?
This is not the way.

Try clicking the ⋯ next to your doc name or using a keyboard shortcut (

CtrlP

) instead.