Image:Overfitting.png

Un article de Wikipédia, l'encyclopédie libre.

Ci-dessous, retrouvez page de description du fichier provenant de Commons.

Overfitting (for example in neural networks training). The red curve is the error on the validation set over several epochs. The blue curve is the error of the training set. When the error for the validation set increases while the training error steadily decreases then a problem of overfitting may occur (the learning is too specialized and does not generalize enough)


GNU head Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation license, Version 1.2 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the section entitled "GNU Free Documentation license".

Aragonés | العربية | Asturianu | Български | বাংলা | ইমার ঠার/বিষ্ণুপ্রিয়া মণিপুরী | Brezhoneg | Bosanski | Català | Cebuano | Česky | Dansk | Deutsch | Ελληνικά | English | Esperanto | Español | Eesti | Euskara | فارسی | Suomi | Français | Gaeilge | Galego | עברית | Hrvatski | Magyar | Bahasa Indonesia | Ido | Íslenska | Italiano | 日本語 | ქართული | ភាសាខ្មែរ | 한국어 | Kurdî / كوردی | Latina | Lëtzebuergesch | Lietuvių | Bahasa Melayu | Nnapulitano | Nederlands | ‪Norsk (nynorsk)‬ | ‪Norsk (bokmål)‬ | Occitan | Polski | Português | Română | Русский | Slovenčina | Slovenščina | Shqip | Српски / Srpski | Svenska | తెలుగు | ไทย | Türkçe | Українська | اردو | Tiếng Việt | Volapük | Yorùbá | ‪中文(中国大陆)‬ | ‪中文(台灣)‬ | +/-

Historique du fichier

Cliquer sur une date et une heure pour voir le fichier tel qu’il était à ce moment-là

Date et heureDimensionsUtilisateurCommentaire
actuel11 août 2005 à 01:231 220×900 (45 Kio)Dake (Overfitting (for example in neural networks training). The red curve is the error on the validation set over several epochs. The blue curve is the error of the training set. When the error for the validation set increases while the training error steadily)

La page ci-dessous contient cette image :