A robust compile time method for scheduling task parallelism on distributed memory machines.
Sekhar DarbhaSantosh PandePublished in: IEEE PACT (1996)
Keyphrases
- high accuracy
- probabilistic model
- high precision
- detection method
- preprocessing
- significant improvement
- cost function
- classification method
- classification accuracy
- clustering method
- theoretical analysis
- feature set
- computational cost
- computational complexity
- fully automatic
- parallel processing
- optimization method
- neural network
- synthetic data
- optimization algorithm
- image registration
- scheduling problem
- dynamic programming
- multiscale