An Experimental Study of SYCL Task Graph Parallelism for Large-Scale Machine Learning Workloads.
Cheng-Hsiang ChiuDian-Lun LinTsung-Wei HuangPublished in: Euro-Par Workshops (2021)
Keyphrases
- machine learning
- machine learning methods
- graph theory
- data mining
- real life
- weighted graph
- graph representation
- computer vision
- feature selection
- knowledge acquisition
- text mining
- connected components
- parallel processing
- machine learning algorithms
- computer systems
- pattern recognition
- database systems
- parallel computation
- graph mining
- graph model
- graph structure
- shared memory
- real world
- random walk
- model selection
- supervised learning
- information extraction
- data analysis
- databases
- structured data
- inductive learning
- decision trees
- spanning tree
- undirected graph
- computational power
- artificial intelligence
- graph based algorithm
- massive graphs