A Systematic Evaluation of Recurrent Neural Network Models for Edge Intelligence and Human Activity Recognition Applications

A Systematic Evaluation of Recurrent Neural Network Models for Edge Intelligence and Human Activity Recognition Applications

The Recurrent Neural Networks (RNNs) are an essential class of supervised learning algorithms. Complex tasks like speech recognition, machine translation, sentiment classification, weather prediction, etc., are now performed by well-trained RNNs. Local or cloud-based GPU machines are used to train t...

Full description

Bibliographic Details
Main Authors: Varsha S. Lalapura, Veerender Reddy Bhimavarapu, J. Amudha, Hariram Selvamurugan Satheesh
Format: Article
Language:English
Published: MDPI AG 2024-02-01
Series:Algorithms
Subjects:
Recurrent Neural Networks
hyperparameter tuning
compression
weights and biases (wandb)
sparsity
low rank
Online Access:https://www.mdpi.com/1999-4893/17/3/104

Internet

https://www.mdpi.com/1999-4893/17/3/104

Similar Items