Page 232 - AI Computer 10

2. A hospital uses an AI model to detect a rare disease. After testing the model on various case histories of
patients, the model produced the output, which has been displayed in the confusion matrix given below:

Actual Positive Actual Negative
Predicted Positive 30 10

Predicted Negative 20 940
On the basis of the above data, answer the following questions:
a. Compute the F1-score for detecting the disease.

b. Compute the Precision for the above data.
c. Why is precision important in this case?

3. A self-driving car uses AI to detect pedestrians. It sometimes fails to detect pedestrians crossing the road
and takes them for objects, thus creating hazardous situations. The developers of the AI system wish to
evaluate the system’s performance and improve it
a. Which evaluation metric should be prioritised?
b. What could be the consequences of a high False Negatives in this scenario?

c. Describe any one factor that could have led this error in the car’s AI model.

98
98