Zobrazeno 1 - 5
of 5
pro vyhledávání: '"Yilmaz, Fatih Furkan"'
Test-time Recalibration of Conformal Predictors Under Distribution Shift Based on Unlabeled Examples
Autor:
Yilmaz, Fatih Furkan, Heckel, Reinhard
Modern image classifiers are very accurate, but the predictions come without uncertainty estimates. Conformal predictors provide uncertainty estimates by computing a set of classes containing the correct class with a user-specified probability based
Externí odkaz:
http://arxiv.org/abs/2210.04166
Autor:
Yilmaz, Fatih Furkan, Heckel, Reinhard
The risk of overparameterized models, in particular deep neural networks, is often double-descent shaped as a function of the model size. Recently, it was shown that the risk as a function of the early-stopping time can also be double-descent shaped,
Externí odkaz:
http://arxiv.org/abs/2206.01378
Autor:
Heckel, Reinhard, Yilmaz, Fatih Furkan
Over-parameterized models, such as large deep networks, often exhibit a double descent phenomenon, whereas a function of model size, error first decreases, increases, and decreases at last. This intriguing double descent behavior also occurs as a fun
Externí odkaz:
http://arxiv.org/abs/2007.10099
Autor:
Yilmaz, Fatih Furkan, Heckel, Reinhard
Image classification problems are typically addressed by first collecting examples with candidate labels, second cleaning the candidate labels manually, and third training a deep neural network on the clean examples. The manual labeling step is often
Externí odkaz:
http://arxiv.org/abs/1910.09055
Autor:
Yilmaz, Fatih Furkan
Bu çalışmada Türkiye'nin ilk ve tek doğal soda külü üreticisi olan Eti Soda AŞ'nin nihai atıklarının içerdiği değerli sodyum karbonat (Na2CO3) minerallerinin sodyum klorür safsızlıklarından arındırılarak kristallendirilmesi ama
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=od_____10208::b06bb3c99391cce9fb2d35cac747e1d1
https://acikbilim.yok.gov.tr/handle/20.500.12812/298081
https://acikbilim.yok.gov.tr/handle/20.500.12812/298081