Toward principled regularization of deep networks-From weight decay to feature contraction.
basic_science · Level V
Where this comes from
- Record sourced from PubMed, PMID 33137722.
- Also identified by DOI 10.1126/scirobotics.aaw1329.
- No licence information is recorded for this record.
- Because redistribution is not established, this page shows the abstract only. Follow the links below for the full text.
Abstract
Training deep artificial neural networks for classification problems may benefit from exploiting intrinsic class similarities by way of network regularization that compensates for a drawback in the commonly used target error.