Your manager wants to use the total accurate classification rate (percent of all cases properly classified) as the metric to evaluate the division's models. Is this a good idea? Why or why not? Select all that apply. A. Good idea; There is no difference between a false positive and a false negative error. A percent of all cases properly classified separates correct classifications from errors. B. Not a good idea; There are frequently differential costs to errors. One error may have larger consequences than another so a percent of correct classifications would not account for these varying costs. C. Not a good idea; We are frequently predicting classification in which the probability of each group is quite different, simply guessing the majority category will frequently result in an excellent overall classification rate. D. Not a good idea; The division's models always results in a 95% accuracy rate. Using the total accurate classification rate would result in all models appearing equal when they are not