L2 regularization can be expected to give superior performance over L1.
Does "performance" here refer to training performance or overall accuracy of the model?
1 Matching Annotations
- Sep 2017
-
cs231n.github.io cs231n.github.io
Tags
Annotators
URL
-