Page 233 - AI Ver 3.0 class 10_Flipbook
P. 233

So, the final Confusion matrix will be as follows:


                                                                          Predicted
                                                  Confusion Matrix
                                                                        Yes       No

                                                              Yes        3         2
                                                   Actual
                                                              No         2         3


                 Here, the total number of correct predictions are 6 out of 10
                 The classification outcomes based on the different values of actual and predicted labels are as follows:
                    • True Positive
                    • True Negative

                    • False Positive
                    • False Negative

                 True Positive

                 A True Positive occurs when a model correctly predicts a positive outcome. In the above example, the value of the
                 True Positive is depicted in the shaded region.


                                                                          Predicted
                                                  Confusion Matrix
                                                                        Yes       No

                                                              Yes        3         2
                                                   Actual
                                                              No         2         3

                 Some more examples of True Positive are as follows:
                    • Medical Diagnosis - A machine learning model predicts whether the patient has asthma.

                   True Positive: The model predicts that the patient actually has asthma.
                    • Face Recognition Security System - A security system identifies individuals who are authorised to enter a
                   restricted area.
                   True Positive: The system recognises an authorized employee correctly.

                 True Negative

                 A  True Negative  occurs  when  a  model  correctly  predicts  a  negative  outcome.  When  the  model’s  negative
                 prediction is same as the actual outcome, it’s the case of True Negative. In the given example the shaded region
                 depicts the True Negative scenario.


                                                                          Predicted
                                                  Confusion Matrix
                                                                        Yes       No
                                                              Yes        3         2
                                                   Actual
                                                              No         2         3

                                                                                           Evaluating Models    231
   228   229   230   231   232   233   234   235   236   237   238