Pages that link to "Item:Q4602826"
From MaRDI portal
The following pages link to Riemannian metrics for neural networks I: feedforward networks (Q4602826):
Displaying 13 items.
- Online natural gradient as a Kalman filter (Q1786581) (← links)
- Wasserstein proximal of GANs (Q2117893) (← links)
- Parametrisation independence of the natural gradient in overparametrised systems (Q2117935) (← links)
- Understanding approximate Fisher information for fast convergence of natural gradient descent in wide neural networks* (Q5020052) (← links)
- (Q5148924) (← links)
- Dynamics of Learning in MLP: Natural Gradient and Singularity Revisited (Q5157104) (← links)
- Adaptive Learning Algorithm Convergence in Passive and Reactive Environments (Q5157257) (← links)
- Stochastic sub-sampled Newton method with variance reduction (Q5204645) (← links)
- Universal statistics of Fisher information in deep neural networks: mean field approach<sup>*</sup> (Q5857452) (← links)
- On the locality of the natural gradient for learning in deep Bayesian networks (Q6115102) (← links)
- Invariance properties of the natural gradient in overparametrised systems (Q6115103) (← links)
- Efficient approximations of the fisher matrix in neural networks using kronecker product singular value decomposition (Q6127062) (← links)
- Natural reweighted wake-sleep (Q6488734) (← links)
