skada.deep.DANLoss

class skada.deep.DANLoss(sigmas=None)

Loss DAN

This loss reduces the MMD distance between source features and target features. From [14].

Parameters:

sigmasarray-like, optional (default=None)

The sigmas for the Gaussian kernel.

References

[14]

Mingsheng Long et. al. Learning Transferable Features with Deep Adaptation Networks. In ICML, 2015.

forward(y_s, y_pred_s, y_pred_t, domain_pred_s, domain_pred_t, features_s, features_t)

Compute the domain adaptation loss