In some predictive modelling projects, we may have variables that most of the observations have the same value, while the small percentage rest ones are populated with meaningful values. For example, 90% observations have values=0 but the rest 10% ha…