LSTM Recurrent Neural Network (RNN) | Explained in Detail

LSTM Recurrent Neural Network is a special version of the RNN model. It stands for Long Short-Term Memory. The simple RNN has a problem that it cannot remember the context in a long sentence because it quickly loses information. And that is why Simple RNN has an only short-term memory.

LSTM has both long-term and short-term memory. It can store any contextual information for a long time.

LSTM has 2 internal states.
1.) Memory Cell State which acts as a long term memory
2.) Hidden State which acts as a short term memory

The main working components in LSTM are gates. There are 3 types of gates in LSTM:
1.) Forget Gate
2.) Input Gate
3.) Output Gate

In the video, we have understood LSTM Recurrent Neural Network in detail.

➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖

Timestamps:
0:00 Intro
1:36 Problem with RNN
5:30 LSTM Overview
7:42 Forget Gate
10:39 Input Gate
13:39 Equations and other details
16:41 Summary of LSTM
18:23 LSTM through different times
19:01 End

➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖

📕📄 Quiz: forms.gle/no29DhL1pF1dsFw28

➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖

Follow my entire playlist on Recurrent Neural Network (RNN) :

📕 RNN Playlist:    • What is Recurrent Neural Network in Deep L...  

➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖

✔ CNN Playlist:    • What is CNN in deep learning? Convolutiona...  

✔ Complete Neural Network:    • How Neural Networks work in Machine Learni...  

✔ Complete Logistic Regression Playlist:    • Logistic Regression Machine Learning Examp...  

✔ Complete Linear Regression Playlist:    • What is Linear Regression in Machine Learn...  

➖➖➖➖➖➖➖➖➖➖➖➖➖➖➖

If you want to ride on the Lane of Machine Learning, then Subscribe ▶ to my channel here:    / @machinelearningwithjay