Model Evaluations (Precision,Recall, F1 Score) using Stratified K-Fold Cross Validation Machine Learning

Question

I have a data Set on which i have applied Stratified K Fold Cross Validation and split the data into 5 folds. Then i have applied Logistic Regression. For Evaluation i have got precision recall and f1 score for each fold. Finally i have to report these evaluations in numbers (precision,recall and f1 score) am i allowed to average precision for all the 5 folds to present just average value same for recall and f1 score. as i have list of five values for each evaluation score after K Fold.

Answer 1

Yes of course you can calculate the mean to get the mean performance over your 5 folds. You can also get the standard deviation of your 5 values. Applying a 5-fold cross-validation is equivalent to training your Logistic Regression model 5 times, so what you need to think about is what to do with the 5 values you get. Maybe you just want to get the max or min of them. It really depends on what kind of evaluations you want to report, which means it depends on real-life scenarios.

Hope my answer helps.