Oversampling and undersampling in data analysis

Within statistics, oversampling and undersampling in data analysis are techniques used to adjust the class distribution of a data set (i.e. the ratio between the different classes/categories represented).