Page 218 - Ai_417_V3.0_C9_Flipbook
P. 218

ROC Curve

              The Receiver Operating Characteristic (ROC) curve is a graphical representation that illustrates the performance of
              a binary classifier system at varying threshold values. It plots the True Positive Rate (TPR) against the False Positive
              Rate (FPR) at various threshold settings.
              This curve plots two parameters:
                 • True Positive Rate (Sensitivity) is the proportion of actual positive cases that are correctly identified by the
                classifier.
                                                                    TP
                                                            TPR =
                                                                  TP + FN
                 • False Positive Rate is the proportion of actual negative cases that are incorrectly classified as positive.
                                                                    FP
                                                            FPR =
                                                                  FP + TN
              To generate an ROC curve, you need to perform the following tasks:
                 • Vary the threshold of your classifier, usually ranging from 0 to 1, and calculate TPR and FPR at each threshold.
                 • Plot these TPR and FPR values on a graph. TPR is plotted on the y-axis, and FPR is plotted on the x-axis.

              Lowering the classification threshold classifies more items as positive, thus increasing both false positives and true
              positives. The following figure shows a typical ROC curve.


                                                                    TP vs FP rate at one
                                                                    decision threshold
                                                      TP vs FP rate at
                                                      another decision
                                                  1   threshold





                                                   TP Rate







                                                  0
                                                    0             FP Rate       1




              Evaluate: Exoplanet Use Case
              At this particular stage, we may need to
              evaluate the model to find out the accuracy of
              the model that makes the best prediction.
              ROC  is  a  metric  that  is  used  to  find  out  the
              accuracy of the model.

              Figure shows the accuracy of the model using
              ROC curve.




                    216     Touchpad Artificial Intelligence (Ver. 3.0)-IX
   213   214   215   216   217   218   219   220   221   222   223