Neural Tangent Kernel

نواة المماس العصبي

نواة تصف سلوك الشبكات العصبية لانهائية العرض أثناء التدريب بالانحدار التدريجي، وتربط نظرياً بين أساليب النواة والشبكات العصبية العميقة. بُنيت على إطار السمات العشوائية.

A kernel that describes the behavior of infinitely wide neural networks during gradient descent training, theoretically connecting kernel methods to deep neural networks. It builds directly on the random features framework.

Also translated asنواة الخط المماس العصبية، NTK

First appears in this corpus in: Neural Tangent Kernel: Convergence and Generalization in Neural Networks (2018)

Appears in these papers

Reconciling Modern Machine-Learning Practice and the Classical Bias–Variance Trade-Off2019 in the sky ✦
Tensor Programs V: Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer2022 in the sky ✦
Tensor Programs V: Tuning Large Neural Networks via Zero-Shot Hyperparameter Transfer2022 in the sky ✦
Neural Tangent Kernel: Convergence and Generalization in Neural Networks2018 in the sky ✦