GRU
الوحدة العودية البوابية
خلية شبكة عصبية عودية اقترحها تشو وزملاؤه عام 2014، تستخدم بوابتَي تحديث وإعادة تعيين للتحكم في تدفق المعلومات عبر الزمن، بأداء مقارب لخلية LSTM لكن بمعاملات أقل وحساب أسرع.
A recurrent neural network cell proposed by Cho et al. in 2014 that uses update and reset gates to control information flow across time, achieving performance comparable to the LSTM cell with fewer parameters and faster computation.
Also translated asGRU، Gated Recurrent Unit، وحدة عودية ذات بوابات
First appears in this corpus in: Learning Long-Term Dependencies with Gradient Descent is Difficult (1994)
Appears in these papers
- Learning Long-Term Dependencies with Gradient Descent is Difficult1994in the sky ✦
- Representation Learning with Contrastive Predictive Coding2018in the sky ✦
- Deep Speech 2: End-to-End Speech Recognition in English and Mandarin2015in the sky ✦
- Mastering Diverse Domains Through World Models2023in the sky ✦
- Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling2014in the sky ✦
- Learning Phrase Representations Using RNN Encoder-Decoder for Statistical Machine Translation2014in the sky ✦