Question 16
Domain 2 — AI Operations, Lifecycle, and Control EnvironmentWhich AI model type is most appropriate for generating realistic synthetic data that mimics training distributions?
Correct answer: C
Explanation
Generative adversarial networks (GANs) are designed to learn a data distribution and produce new samples that resemble the training set. They use a generator and discriminator in competition, which helps the model create realistic synthetic data that mimics the original distribution.
Why each option is right or wrong
A. Convolutional neural networks (CNNs)
B. Recurrent neural networks (RNNs)
C. Generative adversarial networks (GANs)
GANs are the standard choice when the task is to learn an underlying data distribution and then sample new records that look statistically similar to the training set. In the original GAN formulation by Goodfellow et al. (2014), the generator is trained against a discriminator in a minimax objective, which specifically drives the model toward producing high-fidelity synthetic examples rather than merely classifying or predicting labels.
D. Decision trees