You are given a task to evaluate how well a new fire mapping algorithm works. The fire mapping algorithm is a Bayesian classifier which labels all the locations into two classes, burned and unburned. To evaluate the algorithm, two regions are tested. The confusion matrices of these two regions are given in Table 1 and Table 2 Predicted class Data set 1 Burned 30 10 Unburned 20 40 Actual Burned Class Unburned Predicted class Data set'2 Burned 30 1000 Unburned 20 4000 ActualBuned ClassUnburned a) Calculate the TNR, FPR, Precision and Recall of M for the "burned" class for both these data sets b) Is there a difference in their values for the two data sets? If so, what characteristic of the data sets (that are used to derive the above contingency tables) lead to the differences between the values of (TPR.FPR) and (Precision, Recall) that you observe above c) Compute Accuracy and F-measure with respect to 'burned' class for dataset 2. You are given a task to evaluate how well a new fire mapping algorithm works. The fire mapping algorithm is a Bayesian classifier which labels all the locations into two classes, burned and unburned. To evaluate the algorithm, two regions are tested. The confusion matrices of these two regions are given in Table 1 and Table 2 Predicted class Data set 1 Burned 30 10 Unburned 20 40 Actual Burned Class Unburned Predicted class Data set'2 Burned 30 1000 Unburned 20 4000 ActualBuned ClassUnburned a) Calculate the TNR, FPR, Precision and Recall of M for the "burned" class for both these data sets b) Is there a difference in their values for the two data sets? If so, what characteristic of the data sets (that are used to derive the above contingency tables) lead to the differences between the values of (TPR.FPR) and (Precision, Recall) that you observe above c) Compute Accuracy and F-measure with respect to 'burned' class for dataset 2