ASHL: An Adaptive Multi-Stage Distributed Deep Learning Training Scheme for Heterogeneous Environments.
Zhaoyan ShenQingxiang TangTianren ZhouYuhao ZhangZhiping JiaDongxiao YuZhiyong ZhangBingzhe LiPublished in: IEEE Trans. Computers (2024)
Keyphrases
- multistage
- heterogeneous environments
- deep learning
- deep architectures
- restricted boltzmann machine
- highly distributed
- single stage
- web services
- unsupervised learning
- dynamic programming
- unsupervised feature learning
- machine learning
- deep belief networks
- mental models
- lot sizing
- optimal policy
- supervised learning
- training set
- training examples
- distributed environment
- online learning
- object recognition
- fault tolerant
- weakly supervised
- distributed systems
- multiscale