Page 238 - AI Ver 3.0 class 10_Flipbook
P. 238

Where should we use Precision?

              Precision is generally used for unbalanced datasets, when dealing with the False Positives become important, and
              the model needs to reduce the FPs as much as possible.
              Precision use case examples:
                 • In email spam detection, False Positives can be costly, leading to missed business opportunities or communication
                 breakdowns.
                 • Precision is critical when the cost of a False Positive is high, such as in fraud detection systems where unnecessary
                 transaction rejections can cause inconvenience to customers.


                            Task                                                          21 st  Century   #Critical Thinking
                                                                                              Skills


                 Now, find out is good Precision equivalent to a good model performance?









                       Recall


              The recall is the measure of our model correctly identifying True Positives. Recall is defined as the fraction of
              correctly predicted Positive cases to Total Actual Positive cases. It majorly takes into account the True Actual cases
              i.e. it is a measure of our model correctly identifying True Positives.
                                                      No. of correct positive predictions
                                             Recall =
                                                         Total actual positive values
                                                        TP
                                                   =
                                                      TP+FN

              Where should we use Recall?

              Recall is generally used for unbalanced dataset, when dealing with the False Negatives become important and the
              model needs to reduce the FNs as much as possible.

              Recall use case examples:
                 • Fraud Detection system detects fraudulent transactions in a financial institution. A False Negative would allow
                 a fraudulent transaction to go undetected, potentially resulting in significant financial loss.
                 • Fire Alarm system alerts users of potential fires in a building. A False Negative means the system fails to detect
                 an actual fire, leading to catastrophic consequences.
              Which Metric is more Important—Recall or Precision?


                 • Cases where cost of False Negative > False Positive:
                 ○   Can be riskier and more unfavourable, as nobody expected snowfall and there was no Prediction also but still
                    snowfall was there.
                 ○   Can be costly and harmful, as when no rain is predicted and the farmers did not start the harvest as they
                    thought that since no rain is predicted so their wheat crop which is ready to harvest can be done in a day or
                    two. The model that was supposed to predict the rain but did not do so spoiled the whole crop.

                    236     Touchpad Artificial Intelligence (Ver. 3.0)-X
   233   234   235   236   237   238   239   240   241   242   243