Page 24 - CT_AI_Class-6
P. 24

MACHINE LEARNING

                  Machine Learning is a part of Artificial Intelligence that allows computers to learn from data
                  without being directly programmed. It studies patterns and relationships in the data to
                  understand how things work. Based on this learning, it can make decisions or predictions. The
                  more data it receives, the better it becomes at performing tasks. It is used in many areas such
                  as recommendations, spam filtering and smart assistants.

                  Machine learning mainly has three types of learning:

                                                                                             fact bits
                                                                                       Arthur Samuel, a pioneer
                                      Machine learning
                                                                                        in the field of artificial
                                                                                      intelligence and computer
                                                                                       gaming, coined the term
                       Supervised        Unsupervised       Reinforcement                Machine Learning.
                         learning          learning            learning






                  Supervised learning

                  Supervised learning is a type of machine learning in which a model is trained using labelled data.
                  In this method, each piece of data is given a tag or label that clearly shows what it represents.
                  These labels help the model learn by guiding it towards the correct answers. A label is simply

                  a piece of information used to identify or classify data. For example, in a class, students are
                  assessed based on their performance in exams and assignments. Their results are then placed
                  into categories such as Outstanding, Very Good, Satisfactory or Needs Improvement.
                  Some simple examples of labelled data:

                    Email spam detection: Sorting emails into categories such as spam and not spam.

                    Image recognition: Identifying objects in pictures, such as cats, dogs or cars.
                    House price prediction: Estimating the price of a house using details like size, location and

                     features.
                    Handwritten digit recognition: Recognising numbers written by hand, such as those on bank

                     cheques.
                    Medical diagnosis: Predicting possible diseases based on patient data and symptoms.
                  Let us understand supervised learning with a simple example.

                  Suppose you want to build a model that can identify the type of fruit using its weight and size. You
                  start with a labelled dataset where each fruit is already identified:
                    Apple   Weight: 200 grams, Size: Medium           Banana   Weight: 120 grams, Size: Long

                    Orange   Weight: 150 grams, Size: Round           Grape   Weight: 5 grams, Size: Small





                   22        Artificial Intelligence (CT & AI)-VI
   19   20   21   22   23   24   25   26   27   28   29