Efficient distributed machine learning for large-scale models by reducing redundant communication.
Harumichi YokoyamaTakuya ArakiPublished in: SmartWorld/SCALCOM/UIC/ATC/CBDCom/IOP/SCI (2017)
Keyphrases
- machine learning
- model selection
- communication cost
- lightweight
- probabilistic model
- data mining
- fully distributed
- communication overhead
- decision trees
- supervised learning
- computer vision
- machine learning methods
- statistical methods
- learning tasks
- computer networks
- spatially distributed
- communication systems
- data driven approaches
- learning models
- small scale
- statistical models
- complex systems
- knowledge representation
- active learning
- query processing
- computer science
- natural language
- learning algorithm