I am performing 10 Fold cross-validation over my entire dataset for my algorithm. Finally, I am showing the mean RMSE along with the standard deviation. The same operation I am doing for the R^2 ( coefficient of determination). For each of the 10 folds, I obtain a p-value signifying whether my data (edited) is significant or not. My question is can I show a mean of all these 10 p-values?
You certainly can do this, but it would be meaningless. Showing them would be a nice demonstration that you did not really understand what CV is good for.
Btw, the p-value is not about the significance of your model. It's about the data used to fit a given model.
Thank you, Dr. Jochen Wilhelm, for your comment. I apologize for my incorrect understanding of the p-value as I am pretty young in statistics as I feel. I have corrected my question, and I thank you for clearing my doubt.
- Badges
- Science topic