I have a dataset and I want to classify it. Is it necessary to run the 10-fold cross validation technique for more than once in order to generalize my accuracy?
You should use 10-fold cross validation to determine the optimal complexity of the classifier. Not to determine the generalization properties (like accuracy).
So yes, you should used some "10-fold cross test" on top of the 10-fold cross validation in order to determine the a generalization properties (like accuracy).
You should use 10-fold cross validation to determine the optimal complexity of the classifier. Not to determine the generalization properties (like accuracy).
So yes, you should used some "10-fold cross test" on top of the 10-fold cross validation in order to determine the a generalization properties (like accuracy).
Running 10 fold more than once will in general give you the same result, and also to find if your classification can be generalized, you might want to try on another dataset, and make sure you have enough data to begin with
- Badges
- Science topic
- Similar topics
- Biological Science
- Kinesiology
- Running