Perspective: new insights from loss function landscapes of neural networks
Perspective: new insights from loss function landscapes of neural networks
About this item
Full title
Author / Creator
Publisher
Bristol: IOP Publishing
Journal title
Language
English
Formats
Publication information
Publisher
Bristol: IOP Publishing
Subjects
More information
Scope and Contents
Contents
We investigate the structure of the loss function landscape for neural networks subject to dataset mislabelling, increased training set diversity, and reduced node connectivity, using various techniques developed for energy landscape exploration. The benchmarking models are classification problems for atomic geometry optimisation and hand-written d...
Alternative Titles
Full title
Perspective: new insights from loss function landscapes of neural networks
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_hal_primary_oai_HAL_hal_02971899v1
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_hal_primary_oai_HAL_hal_02971899v1
Other Identifiers
ISSN
2632-2153
E-ISSN
2632-2153
DOI
10.1088/2632-2153/ab7aef
