Skip to main content

Table 9 Relative mean standard error (RRSE) values for a different set of workloads and models

From: Runtime prediction of big data jobs: performance comparison of machine learning algorithms and analytical models

Workload f(S) Wordcount linear SVM quadratic Pagerank linear Kmeans linear Graph (NWeight) quadratic
Amdhal equation (1) 0.085 ± 0.005 0.282 ± 0.009 0.113 ± 0.000 0.140 ± 0.019 0.252 ± 0.009
Gustafson equation (2) 0.091 ± 0.004 0.311 ± 0.003 0.120 ± 0.000 0.134 ± 0.009 0.254 ± 0.008
ERNEST equation (3) 0.100 ± 0.006 0.366 ± 0.001 0.127 ± 0.000 0.142 ± 0.017 0.231 ± 0.009
2D plate equation (4) 0.086 ± 0.005 0.272 ± 0.009 0.113 ± 0.000 0.141 ± 0.019 0.226 ± 0.007
Connected graph equation (5) 0.091 ± 0.009 0.268 ± 0.009 0.116 ± 0.000 0.141 ± 0.018 0.244 ± 0.008
Con. graph \(c=1\) equation (6) 0.091 ± 0.009 0.273 ± 0.008 0.118 ± 0.000 0.140 ± 0.018 0.243 ± 0.008
Kernel ridge regression 0.203 ± 0.009 0.234 ± 0.002 0.173 ± 0.001 0.207 ± 0.024 0.178 ± 0.003
Gradient boost regression 0.058 ± 0.003 0.064 ± 0.010 0.033 ± 0.002 0.062 ± 0.038 0.087 ± 0.004
  1. The bold data in each column indicates the smallest RRSE value in the corresponding column