* [[https://en.wikipedia.org/wiki/Keras|Keras]] : Designed to enable fast experimentation with deep neural networks. Keras is built on top of TensorFlow
\\
\\
* How the weights are calculated/readjusted:
* How the weights are calculated/readjusted:
* Backward propagation: If the output is wrong, the system needs to go back and revise its weights
* Backward propagation: If the output is wrong, the system needs to go back and revise its weights
-
* Gradient Descent: TODO
+
* Gradient Descent: See Thomas Nield live course in Oreilly
network_stuff/machine_learning/nns.1685832446.txt.gz · Last modified: (external edit)