Suppose that we apply bagging and random forests respectively to Weekly data set (the same one used in the assignment) and obtain two models M1 and M2. Given the codes as below, which model is from bagging and which one is from random forests? In general, which model (M1 or M2) is more likely to have a higher prediction accuracy on the new testing data set? M1 <- randomForest (Direction M2 <- random Forest (Direction ., data = Weekly, mtry - 2) ., data = Weekly, mtry = 6)