SparCML: high-performance sparse communication for machine learning.
Cédric RenggliSaleh AshkboosMehdi AghagolzadehDan AlistarhTorsten HoeflerPublished in: SC (2019)
Keyphrases
- machine learning
- sparse principal component analysis
- data mining
- machine learning methods
- natural language processing
- machine learning algorithms
- text classification
- high dimensional
- sparse data
- pattern recognition
- feature selection
- communication systems
- active learning
- information extraction
- decision trees
- statistical learning
- computer vision
- learning problems
- learning algorithm
- computer science
- learning systems
- cost effective
- inductive learning
- kernel methods
- machine learning approaches
- semi supervised learning
- information exchange
- communication overhead
- statistical methods
- data acquisition
- sparse representation
- knowledge acquisition
- computational intelligence
- support vector machine
- data analysis
- reinforcement learning
- artificial intelligence