Large width limits of neural networks (original) (raw)
Artificial neural networks are a class of models used in machine learning, and inspired by biological neural networks. They are the core component of modern deep learning algorithms. Computation in artificial neural networks is usually organized into sequential layers of artificial neurons. The number of neurons in a layer is called the layer width. Theoretical analysis of artificial neural networks sometimes considers the limiting case that layer width becomes large or infinite. This limit enables simple analytic statements to be made about neural network predictions, training dynamics, generalization, and loss surfaces. This wide layer limit is also of practical interest, since finite width neural networks often perform strictly better as layer width is increased.
| Property | Value |
|---|---|
| dbo:abstract | Artificial neural networks are a class of models used in machine learning, and inspired by biological neural networks. They are the core component of modern deep learning algorithms. Computation in artificial neural networks is usually organized into sequential layers of artificial neurons. The number of neurons in a layer is called the layer width. Theoretical analysis of artificial neural networks sometimes considers the limiting case that layer width becomes large or infinite. This limit enables simple analytic statements to be made about neural network predictions, training dynamics, generalization, and loss surfaces. This wide layer limit is also of practical interest, since finite width neural networks often perform strictly better as layer width is increased. (en) |
| dbo:wikiPageID | 64415394 (xsd:integer) |
| dbo:wikiPageLength | 8586 (xsd:nonNegativeInteger) |
| dbo:wikiPageRevisionID | 1121159503 (xsd:integer) |
| dbo:wikiPageWikiLink | dbr:Neural_tangent_kernel dbr:Deep_learning dbr:Machine_learning dbc:Machine_learning dbr:Artificial_neural_network dbc:Artificial_neural_networks dbr:Artificial_neuron dbc:Deep_learning dbr:Neural_Network_Gaussian_Process dbr:Neural_circuit dbr:Deep_information_propagation dbr:File:Infinitely_wide_neural_network.webm |
| dbp:wikiPageUsesTemplate | dbt:Reflist |
| dct:subject | dbc:Machine_learning dbc:Artificial_neural_networks dbc:Deep_learning |
| rdfs:comment | Artificial neural networks are a class of models used in machine learning, and inspired by biological neural networks. They are the core component of modern deep learning algorithms. Computation in artificial neural networks is usually organized into sequential layers of artificial neurons. The number of neurons in a layer is called the layer width. Theoretical analysis of artificial neural networks sometimes considers the limiting case that layer width becomes large or infinite. This limit enables simple analytic statements to be made about neural network predictions, training dynamics, generalization, and loss surfaces. This wide layer limit is also of practical interest, since finite width neural networks often perform strictly better as layer width is increased. (en) |
| rdfs:label | Large width limits of neural networks (en) |
| owl:sameAs | wikidata:Large width limits of neural networks dbpedia-fa:Large width limits of neural networks https://global.dbpedia.org/id/DAfqU |
| prov:wasDerivedFrom | wikipedia-en:Large_width_limits_of_neural_networks?oldid=1121159503&ns=0 |
| foaf:isPrimaryTopicOf | wikipedia-en:Large_width_limits_of_neural_networks |
| is dbo:wikiPageWikiLink of | dbr:Neural_tangent_kernel dbr:Artificial_neural_network dbr:Neural_network_Gaussian_process |
| is foaf:primaryTopic of | wikipedia-en:Large_width_limits_of_neural_networks |