Long Short-Term Memory (LSTM)
Long short-term memory is a type of recurrent neural network with internal gates that control what to remember and forget, allowing it to learn longer-range patterns in sequences.
What is Long Short-Term Memory (LSTM)?
Introduced in 1997, LSTMs powered major advances in speech recognition, handwriting recognition and machine translation in the 2010s and remain used for time-series forecasting.
How it works
Input, forget and output gates regulate a cell state that carries information across many time steps.
← All Artificial Intelligence & Superintelligence definitions · ← Back to the A–Z definitions