Parametrisation independence of the natural gradient in overparametrised systems

DOI10.1007/978-3-030-80209-7_78OpenAlexW3181803158MaRDI QIDQ2117935

Nihat Ay, Jesse van Oostrum

Publication date: 22 March 2022

Full work available at URL: https://doi.org/10.1007/978-3-030-80209-7_78

zbMATH Keywords

Riemannian metricinformation geometrynatural gradientdeep learning

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Statistical aspects of information-theoretic topics (62B10)

Related Items (1)

Invariance properties of the natural gradient in overparametrised systems

Cites Work

• Unnamed Item
• Algebraic Geometry and Statistical Learning Theory
• Riemannian metrics for neural networks I: feedforward networks
• On the locality of the natural gradient for learning in deep Bayesian networks

This page was built for publication: Parametrisation independence of the natural gradient in overparametrised systems