Weight initialization

تهيئة الأوزان

إعطاء قيم أولية عشوائية وصغيرة لروابط الشبكة العصبية قبل بدء التدريب. تضمن هذه التقنية بدء التعلم بشكل مستقر وتجنب مشاكل تلاشي أو انفجار التدرج التي تعيق تقدم الخوارزميات.

Also translated asتهيئة القيمة الأولية، تهيئة أوزان

First appears in this corpus in: Foundations of the Theory of Probability (1933)

Appears in these papers

Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification2015 in the sky ✦
Foundations of the Theory of Probability1933 in the sky ✦
Visualizing the Loss Landscape of Neural Nets2018 in the sky ✦
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks2019 in the sky ✦
The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks2019 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 ✦
OPT: Open Pre-Trained Transformer Language Models2022 in the sky ✦
Progressive Growing of GANs for Improved Quality, Stability, and Variation2018 in the sky ✦
Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity2022 in the sky ✦
Very Deep Convolutional Networks for Large-Scale Image Recognition2014 in the sky ✦
Very Deep Convolutional Networks for Large-Scale Image Recognition2014 in the sky ✦