When we learn the parameters of a predictive function and evaluate it on the same data, we make an error: A model that repeats the labels of the samples it's just seen would have a perfect score, but would not be able to predict anything useful on yet-unseen data. Overfitting is the term used to describe this situation. As a rule of thumb, when doing a (supervised) machine learning experiment, a portion of the available data is set aside as a test set X, y.

You can get implementing process by using Python or MATLAB through -

https://scikit-learn.org/stable/modules/cross_validation.html

https://www.mathworks.com/help/fuzzy/fis-parameter-optimization-with-k-fold-cross-validation.html

Mohammad Nasfikur Rahman Khan

Thanks a heap brother, but I agin reiterate that using this K-Fold Cross Validation, how can I find a parameter specific to SVR, called as Tolerance Parameter?

I am not sure about that. I am recommending it for a better reach.