Page 317 - AI Ver 1.0 Class 10
P. 317

Task


                   Now after understanding the concept of Accuracy, according to you how much percentage of accuracy is reasonable
                   to show good performance?

                   _______________________________________________________________________________________________________________________________

                   _______________________________________________________________________________________________________________________________

                   _______________________________________________________________________________________________________________________________

                   _______________________________________________________________________________________________________________________________

                   _______________________________________________________________________________________________________________________________



                 In the above Scenario of Board Exams prediction:
                 Let’s assume

                 True Positives = 0
                 True Negatives = 90
                 Total cases = 100
                 Therefore:

                 Accuracy becomes: ((90 + 0) / 100) *100 = 90%
                 This is a fairly high accuracy for an AI model.
                 The model predicts that there are NO BOARD EXAMS. But in reality, there are 10% chances of Board exams to
                 be conducted. In this case, for 90 cases the model will be right but for 10 cases in which there are chances of
                 conducting Board Exam but still model predicted NO BOARD EXAMS.
                 But this parameter is useless for us as the actual cases which are 10 in our above example are not taken into
                 account.
                 Hence, there is a need to look for another parameter which takes into account the situation as in the above
                 example.

                 Precision
                 Precision is the percentage of True Positive cases and All Predicted Positive Cases.
                 Its formula is:

                                                                      True Positive
                                                      Precision =                       × 100%
                                                                  All Predicted Positives

                                                                    TP
                                                      Precision =          × 100%
                                                                  TP + FP

                 Where,
                 All Predicted Positive = True Positive + False Positive







                                                                                                Evaluation  315
   312   313   314   315   316   317   318   319   320   321   322