0

How can I use the AUC in a plot_model([model], plot = 'auc') function that matches up to a certain cross-validation trial for a pycaret model (specifically rf)? The plot automatically shows AUC = 0.77 but the mean AUC is 0.83 so I would want to use the trial with the closest AUC value, which would be fold #9 in the image below.

Default AUC plot

Table from create_model()

I tried to find possible reasons for the plot_model showing AUC = 0.77 and notice it's the model in the cross-validation set with the lowest AUC value. I noticed the same trend in another example model found in an online tutorial, so perhpas that's what pycaret is doing.

ga-so
  • 1
  • 2

0 Answers0