Suppose you want to predict the churn rate. Consider the two models M1 and M2. Based on the confusion matrix, M1 has an accuracy of 75% and cost of 4017. Meanwhile, M2 has an accuracy of 80% and cost...

Extracted text: Suppose you want to predict the churn rate. Consider the two models M1 and M2. Based on the confusion matrix, M1 has an accuracy of 75% and cost of 4017. Meanwhile, M2 has an accuracy of 80% and cost of 5236. Which is the better model and why? * It depends on the business case which errors the stakeholders want to minimize and which predictions they want to maximize. M1 is the better model because it has a lower cost meaning the errors are minimized and the accuracy shows that the model was able to capture customers that will not churn. M2 is the better model because it has a O higher accuracy so the model was able to capture all churn customers. None because you need to use ROC O (receiver operating characteristic) to better evaluate the two models.