Skip to main content

Table 1 Results from the independent samples Welch’s t-test for both grids. Shown are the t-statistics and the corresponding degree of freedom (df). Reported p-values were Bonferroni-corrected. Combinations missing in this table were significant with p<.001 for both grids

From: Evaluating different machine learning techniques as surrogate for low voltage grids

   CIGRE LV   LV-rural3  
Model A Model B df t df t
Sim RE LR 10.4 56.81*** 2.0 25.54*
Sim RE RF 16.7 -78.38*** 2.1 -13.03
Sim RC LR 11.2 60.83*** 2.0 26.32*
Sim k-NN 11.2 63.75*** 2.0 26.02*
Sim ANN 9.7 64.04*** 2.0 27.77*
Sim LSTM 10.4 1.38 2.0 27.58*
RE LR RC LR 17.3 11.94*** 3.8 17.25**
RE LR k-NN 17.3 18.78*** 2.2 13.02
RE LR LSTM 18.0 -143.72*** 2.6 53.87**
RC LR k-NN 18.0 6.23*** 2.2 -10.71
RC LR ANN 14.5 2.22 3.9 40.01***
k-NN ANN 14.5 -5.40** 2.3 72.84**
ANN LSTM 16.4 -183.52*** 3.2 -7.44
  1. *** p<.001
  2. ** p<.01
  3. * p<.05