Page 235 - AI Ver 3.0 class 10_Flipbook
P. 235

The confusion matrix with all classification outcomes based on the different values of actual and predicted labels
                 can be presented as follows:


                                                                          Predicted
                                          Confusion Matrix
                                                                    YES               NO

                                                               True Positive     False Negative
                                                     YES      Predicted – Yes    Predicted – No
                                                                Actual - Yes       Actual - Yes
                                         Actual
                                                               False Positive    True Negative
                                                     NO       Predicted – Yes    Predicted – No
                                                                Actual - No        Actual - No


                         Accuracy from Confusion Matrix


                 Classification Accuracy is the percentage of correct predictions out of the total observations made by an AI
                 model. It provides a clear picture of how accurate the predictions are for the given model. A high accuracy score
                 generally indicates good performance, as it accounts for all correctly predicted values. The mathematical formula
                 for classification accuracy is:

                                                                  No. of correct predictions
                                        Classification Accuracy =                           × 100
                                                                   Total no. of predictions
                 Where,

                 Total Correct Predictions = True Positive (TP) + True Negative (TN)
                 Total Predictions = True Positive (TP) + True Negative (TN) + False Positive (FP) + False Negative (FN)
                 A prediction is said to be correct if it matches reality. Here we have two conditions in which the Prediction matches
                 with the Reality, i.e., True Positive and True Negative.
                 For example, in a model of predicting whether the credit card transaction is fraudulent or not, the confusion matrix
                 is as follows:


                                                                          Predicted
                                                  Confusion Matrix
                                                                        Yes       No

                                                              Yes        15        14
                                                   Actual
                                                              No         12        10


                 Total Correct Predictions  = TP+TN
                                        = 15+10
                                        = 25

                 Total Predictions      = TP+TN+FP+FN
                                        = 15+10+12+14
                                        = 51




                                                                                           Evaluating Models    233
   230   231   232   233   234   235   236   237   238   239   240