> What is Cost-sensitive learning method is the perspective of class
> How a Cost-sensitive learning method is good to solve class imbalance
> problems instead of sampling techniques.?
If the selected base classifier can handle instance weights, then the
rows in the dataset get reweighted according to the cost matrix before
training the base classifier. This allows you to put more importance
on instances from your minority class.
If the base classifier shouldn't handle instance weights, then
resampling according to the cost matrix is performed before training
the base classifier.