DeepAI AI Chat

Long Short-Term Memory

‹ Logit
Longitudinal Study ›

LSTMs are a form of recurrent neural network invented in the 1990s by Sepp Hochreiter and Juergen Schmidhuber, and now widely used for image, sound and time series analysis, because they help solve the vanishing gradient problem by using a memory gates.

The world's most comprehensive
data science & artificial intelligence
glossary

Get the week's most
popular data science
research in your inbox -
every Saturday

Research that mentions Long Short-Term Memory

MomentumRNN: Integrating Momentum into Recurrent Neural Networks

06/12/2020 ∙ by Tan M. Nguyen ∙ 62
Compression of Deep Learning Models for Text: A Survey

08/12/2020 ∙ by Manish Gupta ∙ 60
Generative time series models using Neural ODE in Variational Autoencoders

01/12/2022 ∙ by M. L. Garsdal ∙ 55
Predicting Coordinated Actuated Traffic Signal Change Times using LSTM Neural Networks

08/10/2020 ∙ by Seifeldeen Eteifa ∙ 47
Real-Time Lip Sync for Live 2D Animation

10/19/2019 ∙ by Deepali Aneja ∙ 46
COVID-19 Cough Classification using Machine Learning and Global Smartphone Recordings

12/02/2020 ∙ by Madhurananda Pahar ∙ 43
A Three-dimensional Convolutional-Recurrent Network for Convective Storm Nowcasting

10/01/2019 ∙ by Wei Zhang ∙ 42