Page 195 - AI Ver 3.0 class 10_Flipbook
P. 195

onion
                                                                                    ginger

















                 For example, if we want to train a model to identify if an image is of an onion or a ginger, we need to train it with
                 multiple images of both onion and ginger along with their labels. The machine will then classify images on the
                 basis of the labels and predict the correct label for testing data. Classification works on discrete dataset. Following
                 are some examples of classification:
                    • Image Classification: In this scenario the model would be trained on different categories of images. The model
                   would learn the pattern as fed by the training test and would be able to categorise images into predefined
                   categories based on the pixel values of the images. For example, when differenet image of raw, ripe and rotten
                   bananas are fed into the model, the trained classification model would analyse the images and predict the most
                   likely category raw banana, ripe banana or rotten banana.





                     Ripe Banana  Ripe Banana  Raw Banana  Rotten Banana                            Ripe Banana
                                                                  Classification
                                                                     Model
                                                                                             Raw Banana
                           Raw Banana  Ripe Banana  Rotten Banana

                           Rotten Banana  Ripe Banana  Rotten Banana                                Rotten Banana
                                    Input                                                           Output

                    • Email Spam Detection: In this scenario, the model is trained with the words in the email, information of the
                   sender, precedence of links to classify the emails under “Spam” or “Not Spam”. The trained model would then
                   analyse the emails and predict them as “Spam” or “Not Spam”. Similar models are being used to categorise calls
                   and messages as “Spam”.




                                                                                    Inbox
                                        Spam
                                                                                            Not Spam
                                                          Classification
                                                             Model
                                        Spam
                                                                                    spam

                                                                                              Spam
                                       Not Spam
                                        Input                                            Output



                                                                          Advanced Concepts of Modeling in AI   193
   190   191   192   193   194   195   196   197   198   199   200