Requirement 1. James wonders whether he should use new data to train a new model in December2025,just prior to his planned investment on January1,2026.What should he do with the older data?In machine learning models it is often better toMachine learning relies on theA management accountantneed not worry about the additional costs of updating the data set as newer data will always result in a more predictive model.will have to judge the costs and benefits of updating a data set.Part 2Requirement 2. How much of the actual performance of the loans(such as repayments, loan restructuring, delayed repayments, hardship plans, etc.) should James use in classifying loans as default or repay when building his prediction model?Information about the actual performance of a loanshouldwill certainlywill notbe available to James when he is predicting whether that loan will default or repay in the future. Therefore, heshould include this informationshould not include any of this informationin the model. This is necessary to