Page 224 - AI Computer 10

7. Recall is the rati o of True Positi ves to the sum of True Positi ves and False Negati ves.
8. F1 Score is the harmonic mean of Precision and Recall.

9. If Precision is low and Recall is high, the F1 Score is high.
10. Accuracy alone is suffi cient to ensure the model’s performance on unseen data.

Answers
1. F 2. F 3. T 4. F 5. T 6. F
7. T 8. T 9. F 10. F
D. Very short answer type questions.

1. What is the purpose of model evaluati on?
Ans. To understand the reliability of an AI model by using a separate test dataset.
2. Defi ne “Reality” in the context of model evaluati on.

Ans. The real outcome of a situati on or an event for which the predicti on is made.
3. What does a True Positi ve outcome signify in model evaluati on?
Ans. The model predicts an event to occur and the event actually occurs.

4. What is the use of a confusion matrix in model evaluati on?
Ans. It summarises the model’s predicti ons and the actual events for which the predicti ons are made in terms
of parameters such as True Positi ves and Negati ves.
5. How is Precision calculated?

Ans. True Positi ves divided by the sum of True Positi ves and False Positi ves.
6. What does a False Negati ve in model evaluati on indicate?
Ans. The model predicts an outcome/event does not occur but in reality, it does.
7. Defi ne F1 Score.

Ans. The harmonic mean of Precision and Recall.
8. When might accuracy alone be insuffi cient for model evaluati on?

Ans. When there is an imbalance in the testi ng dataset or when evaluati ng on unused data.
9. In case of predicti ng water shortage, what does a True Negati ve represent?
Ans. Model predicts no water shortage, and there is no water shortage in reality.
10. What does Recall measure in model evaluati on?

Ans. The rati o of True Positi ves to the sum of True Positi ves and False Negati ves.
E. Short answer type questions.

1. What is the signifi cance of evaluati ng an AI model?
Ans. Model evaluati on helps assess the reliability and performance of the AI model by using a separate test
dataset.
2. Explain the role of the “Reality” conditi on in model evaluati on.
Ans. “Reality” represents the actual event/result for which a predicti on is made by the AI model.

3. Why is a True Positi ve outcome important in model evaluati on?
Ans. A True Positi ve indicates the model correctly predicted a reality, demonstrati ng the model’s eff ecti veness.

90
90