Skip to main content
. 2017 Jun 5;8:15461. doi: 10.1038/ncomms15461

Table 1. Summary of the results.

Test set (training set) Initial distribution Linear model Quadratic model Support vector regressor Artificial neural network
Mean error of single-pulse photon energy (eV) 5.62 0.29 (0.28) 0.30 (0.24) 0.32 (0.27) 0.30 (0.29)
Shape agreement of single-pulse spectrum 67% 88% (88%) 94% (95%) 95% (95%) 97% (97%)
Mean error of double-pulse delay (fs) 6.82 2.07 (2.04) 1.67 (1.58) 1.67 (1.57) 1.59 (1.52)
Mean error of double-pulse photon energy (eV) Pulse 1: 1.45 0.47 (0.49) 0.49 (0.48) 0.50 (0.48) 0.46 (0.47)
  Pulse 2: 1.03 0.44 (0.44) 0.41 (0.39) 0.41 (0.39) 0.40 (0.40)

Mean absolute error or agreement of the different prediction examples obtained from each of the four models. The first column shows the mean error from the average of the initial distribution. In the case of shape agreement, this value corresponds to the mean agreement between each of the single-shot spectra and the mean spectrum. The values for each of the models correspond to the predictions on the test set, while the numbers in brackets correspond to the training set.