brainpy.losses.l2_loss#

brainpy.losses.l2_loss(predicts, targets)[source]#

Computes the L2 loss.

The 0.5 term is standard in “Pattern Recognition and Machine Learning” by Bishop 1, but not “The Elements of Statistical Learning” by Tibshirani.

Parameters

  • predicts (ArrayType) – A vector of arbitrary shape.

  • targets (ArrayType) – A vector of shape compatible with predictions.

Returns

loss – A scalar value containing the l2 loss.

Return type

float

References

1

Bishop, Christopher M. 2006. Pattern Recognition and Machine Learning.