Adaptive gradient methods: efficient implementation and generalization
2019 Plenary talk - Elad Hazan
in
Workshop: Beyond first order methods in machine learning systems
in
Workshop: Beyond first order methods in machine learning systems
Abstract
Adaptive gradient methods have had a transformative impact in deep learning. We will describe recent theoretical and experimental advances in their understanding, including low-memory adaptive preconditioning, and insights into their generalizaton ability.
Chat is not available.
Successful Page Load