Why the dataset was not balanced after the step loan_data[‘good_bad’] obtained and before the step of Splitting Data? >>>>>>>>>>>>>> So, in the exercise we are splitting unbalanced data (number of zeros & ones are not equal)that will lead to inaccurate model. >>>>>>>>>>>>>>>>>>> This note was from the subject: the Credit Risk Modeling in Python, under the PD model, Data preparation, Dependent Variable. Good/ Bad (Default) Definition. Default and Non-default Accounts. >>>>>>>>>>>>>> The machine will learn better if the dataset was balanced.