I know, I know — yet another guide on LSTMs / RNNs / Keras / whatever. There are SO many guides out there — half of them full of false information, with inconsistent terminology — that I felt frustrated enough to write this one-stop guide + resource directory, even for future reference. This guide aims to be a glossary of technical terms and concepts consistent with Keras and the Deep Learning literature. This article assumes a very basic conceptual familiarity with the concept of Neural Networks in general. If you spot something that’s inconsistent with your understanding, please feel free to drop a comment / correct me!

Content Page

1. RNNs and LSTMs
2. Hidden State vs Cell State
3. General Gate Mechanisms
4. Gate Operation Dimensions & “Hidden Size”
5. “Hidden Layers”
6. Model Complexity
7. Quirks with Keras — Return Sequences? Return States?

#deep-learning #equation #lstm #keras #machine-learning