The use of resampling techniques to overcome imbalance of data on the classification algorithm