I cannot reproduce the results that you have provided. I got 5 very different results for each dataset. For NL dataset I obtained results varied between 0.8529-0.8630 AUC. Did you get the average of 5 different runs? That much standard deviation for each run is not normal I think.
I cannot reproduce the results that you have provided. I got 5 very different results for each dataset. For NL dataset I obtained results varied between 0.8529-0.8630 AUC. Did you get the average of 5 different runs? That much standard deviation for each run is not normal I think.