Concept information
Término preferido
Newton’s method
Definición
- An algorithm used to find the minimum or maximum of a real-valued function in order to solve optimization problems, especially in the training of machine learning models.
Concepto genérico
Etiquetas alternativas
- Newton–Raphson method
Ejemplo
- While the log-likelihood function for ME models in (2) is twice differentiable for large scale problems the evaluation of the Hessian matrix is computationally impractical and Newton's method is not competitive with iterative scaling or first order methods. (Malouf, 2002)
En otras lenguas
-
francés
-
algorithme de Newton
-
méthode de Newton-Raphson
URI
http://data.loterre.fr/ark:/67375/8LP-S8J3RLVM-W
{{label}}
{{#each values }} {{! loop through ConceptPropertyValue objects }}
{{#if prefLabel }}
{{/if}}
{{/each}}
{{#if notation }}{{ notation }} {{/if}}{{ prefLabel }}
{{#ifDifferentLabelLang lang }} ({{ lang }}){{/ifDifferentLabelLang}}
{{#if vocabName }}
{{ vocabName }}
{{/if}}