/LogGOPSC

LogGOPSC is extended from LogGOPS by adding new parameter C. The parameter is predicted through a designed simple BP neural network. These are the training data used in experiment.

LogGOPSC

This is training and evaluating data of LogGOPSC, in paper LogGOPSC: A Parallel Computation Model Extending Network Contention into LogGOPS.
The 3 files in this catalogue are:
Exercise_Train: The complete training dataset in LogGOPSC.
Exercise_Eval: The complete evaluating dataset in LogGOPSC.
Exercise_eval.xlsx: The evaluating dataset which diagramed in paper, 100 examples for each message size in 2MB, 4MB, 8MB, 16MB, 32MB.