Difference between revisions of "Gradient Descent Optimization & Challenges"
(→Gradient Descent - Stochastic (SGD), Batch (BGD) & Mini-Batch) |
(→Vanishing & Exploding Gradients Challenges with Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM)) |
||
| Line 20: | Line 20: | ||
== Vanishing & Exploding Gradients Challenges with Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) == | == Vanishing & Exploding Gradients Challenges with Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) == | ||
[[http://www.youtube.com/results?search_query=Difficult+to+Train+an+RNN+LSTM+vanishing YouTube Search]] | [[http://www.youtube.com/results?search_query=Difficult+to+Train+an+RNN+LSTM+vanishing YouTube Search]] | ||
| + | |||
| + | * [[Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Recurrent Neural Network (RNN)]] | ||
<youtube>Pp4oKq4kCYs</youtube> | <youtube>Pp4oKq4kCYs</youtube> | ||
Revision as of 09:30, 24 June 2018
Gradient Descent - Stochastic (SGD), Batch (BGD) & Mini-Batch
Vanishing & Exploding Gradients Problems
Vanishing & Exploding Gradients Challenges with Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM)
