Page 70 - CT_AI_Class-8
P. 70
5. In what way AI systems can be tested to ensure fairness in their decisions? Provide two methods
that can be used to check for bias in AI.
21 st
think & apply Century #Critical Thinking
Skills
1. When an AI system has mostly seen one type of data, it struggles with others. What type of data makes
AI less accurate and more biased?
2. In a hiring AI system, if most of the past hires were men, the system may favour men. In the AI
recruitment example, which group did the system unfairly favour?
3. To make sure an AI system treats all groups equally, it’s essential to carefully check the data. What
should be ensured to make AI systems fairer?
4. AI systems can make unfair decisions if the data used to train them is unbalanced. What issue occurs
when AI systems favour one group over another due to biased data?
5. AI systems can help with medical diagnoses, but they may be less effective if trained on biased or
limited data. How could this impact healthcare if the AI is not trained on diverse data?
cbse handbook questions
1. For the dog training dataset, record the following:
How many images are included?
How are the images similar?
How are the images different?
2. For the cat training dataset, record the following:
How many images are included?
How are the images similar?
How are the images different?
Once your classifier is finished, test your dataset with the images below. Fill the following table
about your testing dataset:
Image Classification Confidence Score Correct?
68 Artificial Intelligence (CT & AI)-VIII

