Long Short Term Memory
Long Short-Term Memory (LSTM) is a type of recurrent neural network architecture that uses a gated memory cell to selectively retain, forget, and output information over long sequences, mitigating the vanishing gradient problem of vanilla RNNs.
5 resources across 1 library
Glossary Terms(5)
Hidden Markov Model
A Hidden Markov Model (HMM) is a statistical model of a system that transitions between unobserved (hidden) states over time, where each hidden state produces…
Convolutional Neural Network
A Convolutional Neural Network (CNN) is a deep learning architecture that uses learnable convolutional filters to automatically extract spatial features from g…
Recurrent Neural Network
A Recurrent Neural Network (RNN) is a neural network architecture designed for sequential data, where connections loop back so that the network's hidden state…
Long Short-Term Memory (LSTM)
Long Short-Term Memory (LSTM) is a type of recurrent neural network architecture that uses a gated memory cell to selectively retain, forget, and output inform…
Gated Recurrent Unit (GRU)
A Gated Recurrent Unit (GRU) is a simplified variant of the LSTM recurrent neural network architecture that uses fewer gates and merges the cell and hidden sta…