Skip to Main Content
Table 1:
Area under the ROC (AUC) Curve for Discriminating Change Points.
NNGNNGPfRiskKLIEPNNW
Original data *0.810 (0.052) 0.759 (0.051) 0.787 (0.032) 0.793 (0.047) 0.576 (0.055) 0.748 (0.050) 
Gaussian noise 0.550 (0.055) 0.510 (0.082) *0.608 (0.078) 0.543 (0.078) 0.491 (0.061) 0.532 (0.058) 
Poisson noise *0.735 (0.028) 0.709 (0.032) 0.599 (0.060) 0.729 (0.043) 0.511 (0.041) 0.699 (0.036) 
NNGNNGPfRiskKLIEPNNW
Original data *0.810 (0.052) 0.759 (0.051) 0.787 (0.032) 0.793 (0.047) 0.576 (0.055) 0.748 (0.050) 
Gaussian noise 0.550 (0.055) 0.510 (0.082) *0.608 (0.078) 0.543 (0.078) 0.491 (0.061) 0.532 (0.058) 
Poisson noise *0.735 (0.028) 0.709 (0.032) 0.599 (0.060) 0.729 (0.043) 0.511 (0.041) 0.699 (0.036) 

Note: In addition to the original data, data corrupted with gaussian and Poisson noise are tested. The methods with the best accuracy are starred, and those with -values less than 0.05 from the single-sided -test are in bold.

Close Modal

or Create an Account

Close Modal
Close Modal