I'm closely following code from this tutorial for my data and its trianing nicely: https://www.tensorflow.org/tutorials/structured_data/imbalanced_data#class_weights
The only key difference I've made (other than dataset) is that I perform k-fold cross validation on my dataset (k=5), and ideally wish to perform repeats (e.g 30) of each fold.
I am producing the plots in this tutorial (AUPRC being most important to me). However, I am currently making a seperate plot for each fold. I want these to all be on one plot (i.e AUPRC for all folds on one plot). Ideally, this could even be a mean of the repeats for each fold (or maybe a line which is the mean of means with confidece bounds etc).
I'm aware I can store the metrics output by each training run and get a mean, distrubution etc. But not how to create these plots. How could this be implemented?
Here's an example of one of my AUPRC plots:
