Page 189 - AI Ver 3.0 class 10_Flipbook
P. 189

Let’s take an example to understand Supervised Learning:

                 Build a model to predict the type of fruit based on its weight and size.
                 Assume that you have a dataset of fruits with their weights and sizes labelled:

                    • Apple → Weight: 200 grams, Size: Medium
                    • Banana → Weight: 120 grams, Size: Long
                    • Orange → Weight: 150 grams, Size: Round

                    • Grape → Weight: 5 grams, Size: Small
                 Features of the data:

                    • Weight: The mass of the fruit.
                    • Size: The shape or dimension (e.g., medium, long, round, small).
                 The  model  is  trained  with  labelled  data  where  the  weights  and  sizes  are  associated  with  specific  fruit  types.
                 The model learns to recognise patterns and relationships between the features (weight and size) and the label
                 (fruit type). So, if the trained model is given new data (e.g., Weight: 120 grams, Size: Long), it predicts the fruit
                 type (Banana) based on the training it received. This example demonstrates how labelled data (weight, size, and
                 corresponding fruit type) helps the model learn and make predictions for new instances.

                           Labelled Data
                                                                            Prediction                   Output

                                                       AI Model                                            Apple
                          Apple  Banana
                                                                                                           Grapes

                         Orange  Grapes                                                                    Banana

                                                                                                           Orange



                                                      Test Data

                 Unsupervised Learning


                 Unsupervised learning approach works on unlabelled dataset. This means that the data which is fed to the machine
                 is random and there is no knowhow available about it to the model. The machine analyses the data and identifies
                 patterns, structures, or relationships on its own without any guidance. The goal is to group or organise data based
                 on similarities or differences.

                 In this model the major features are identified by the machine, which help the user in understanding the data.
                 For example, in the data of 100 cat images, if you want to understand some pattern in the data, you would need
                 to feed this data into the unsupervised learning model and train the machine. Once trained, the machine would
                 identify patterns in the data. These patterns might already be known to the user, like colour or size, or different
                 features of the cats.
                 Unsupervised learning helps discover hidden patterns in unlabelled data. For example, imagine a photo gallery
                 app  that  automatically  organises  a  user's  photos  based  on  their  content.  The  photos  aren’t  pre-labelled  as
                 "family," "vacation," "pets," or "friends." Instead, the app uses unsupervised learning to analyse the photos and
                 group them based on similarities, such as recognising the same faces, landscapes, or objects like pets or vehicles.

                                                                          Advanced Concepts of Modeling in AI   187
   184   185   186   187   188   189   190   191   192   193   194