Q67 — AWS AIF-C01 Ch.2

Question 67 of 100 | ← Chapter 2

In the ML context, what does bias refer to?

• A. Model architecture complexity
• B. Randomness or noise in training data
• C. Difference between predicted and actual values
• D. Systematic error or tendency of a model to make incorrect predictions for certain data ✓

Correct Answer: D. Systematic error or tendency of a model to make incorrect predictions for certain data

Explanation

In machine learning, bias refers to option D: systematic error or tendency of a model to make incorrect predictions for certain data. A. Model architecture complexity relates more closely to variance—not bias. B. Randomness or noise in training data contributes to variance and overfitting, not systematic bias. C. The difference between predicted and actual values reflects total error, which comprises both bias and variance components—not bias alone. D. Bias represents inherent, systematic deviation due to oversimplified assumptions (e.g., linear model for nonlinear data) or unrepresentative training data—leading to consistent errors across predictions. High bias causes underfitting and poor generalization, even on training data. Reducing bias typically requires richer features, more complex models, or better data representation.