Question

question 2 of 25
which scenario is most likely to produce algorithmic bias?

a. a computer isnt trained with a wide enough variety of individuals of different ages, genders, facial shapes, and skin tones.

b. a computer is programmed with too many images of a wide variety of individuals, causing it to become overwhelmed.

c. a programmer doesnt train a computer how to tell the difference between human beings and inanimate objects.

d. a programmer writes an exceptionally long and overly complicated program.

Explanation:

Algorithmic bias often arises from insufficient or non - representative training data. Option A describes a situation where the computer's training data lacks diversity in terms of age, gender, facial shape, and skin tone. This lack of diverse training can lead the algorithm to be biased towards the groups that are over - represented in the training data and under - perform or misclassify for the under - represented groups. Option B's "overwhelmed" computer due to too many diverse images is not a typical cause of algorithmic bias. Option C is about the computer not being able to distinguish between humans and inanimate objects, which is a different kind of problem (classification of object types) rather than bias. Option D's long and complicated program may lead to inefficiencies or errors in the program's execution, but not necessarily algorithmic bias.

Answer:

A. A computer isn't trained with a wide enough variety of individuals of different ages, genders, facial shapes, and skin tones.