Q48 — AWS AIF-C01 Ch.1

Question 48 of 100 | ← Chapter 1

An AI developer is building a model to generate diverse human-like images. The AI developer discovers bias in the input data, and this bias affects image generation and propagates into the model. Which technique can address this issue?

• A. Data augmentation for imbalanced classes ✓
• B. Model monitoring for class distribution
• C. Retrieval-Augmented Generation (RAG)
• D. Image watermark detection

Correct Answer: A. Data augmentation for imbalanced classes

Explanation

This question tests knowledge of techniques to mitigate input data bias in AI models. When training data exhibits bias (e.g., underrepresentation of certain demographics), data augmentation for imbalanced classes helps balance the dataset by synthetically increasing minority samples or applying transformations—reducing bias propagation into generated outputs. Option B (monitoring) detects but does not resolve bias. Option C (RAG) enhances factual grounding in LLMs but is irrelevant to image generation bias. Option D (watermark detection) addresses provenance/security, not bias mitigation. Thus, option A is the correct technique.