Page 23 - CT_AI_Class-8
P. 23

Student       Gadget         Screen         Sleep        Outdoor      Academic        Screen
                                    Type          Time          Hours        Activity        Score      Time Level

                   Student 5          1             4              6            60             75             1


                 During training, the AI model studies the data and finds patterns. For example, the model may
                 learn:

                    Students who use mobile phones for longer hours are more likely to have high screen time.

                    Students who get less sleep tend to spend more time on gadgets.
                    Students who spend less time in outdoor activities usually use gadgets more.

                    Higher screen time may lead to lower academic performance.

                    Students  who have adequate  sleep  and  limited  screen  time  generally  perform better
                    academically.

                    Increased outdoor activities usually help in reducing gadget usage.
                    Balanced routine with study, sleep and outdoor activity leads to healthier screen time
                    habits.

                 These patterns help the AI model understand relationships between different factors and make
                 predictions about students' gadget usage. This learning process is called model training.


                 Model Evaluation and Refinement

                 After completing Model Development and Training, the next stage in the AI Project Life Cycle
                 is  Model  Evaluation  and Refinement.  In this  stage,  the  trained  model  is  tested  to  check  how
                 accurately it performs and whether it can make reliable predictions.

                 Model evaluation helps us understand:

                    How well the model has learned

                    Whether the model is making correct predictions

                    Whether improvements are required
                 The model is tested using new data known as test data. This data was not used during training, so

                 it helps in checking how well the model works in real-life situations.

                 Suppose, you teach your friend to identify students with high screen time using some examples.
                 After teaching, you give them new student records. If your friend correctly identifies most of them,
                 it means they learned properly. This is similar to model evaluation.

                 From the Gadget Screen Time data, the AI model may learn patterns such as:

                    Students who use mobile phones for longer hours are more likely to have high screen time.
                    Students who get less sleep tend to spend more time on gadgets.





                                                                                           AI Project Lifecycle    21
   18   19   20   21   22   23   24   25   26   27   28