Analysis of Gradient Vanishing of RNNs and Performance Comparison
A recurrent neural network (RNN) combines variable-length input data with a hidden state that depends on previous time steps to generate output data. RNNs have been widely used in time-series data analysis, and various RNN algorithms have been proposed, such as the standard RNN, long short-term memo...
Saved in:
| Main Author: | |
|---|---|
| Format: | article |
| Language: | EN |
| Published: |
MDPI AG
2021
|
| Subjects: | |
| Online Access: | https://doaj.org/article/c707bedefb6643e791954367081254ca |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|