Changing one categorical attribute into a series of binary attributes is known as ________.

• A. One-Hot Encoding
• B. Normalization
• C. Binarization
• D. Dummy Coding

Answer: (D)

Transforming a categorical attribute into a set of binary indicators is known as dummy coding. This approach represents each category (often using k-1 binary columns with a reference category left as zeros) so models that require numeric inputs can use the categorical information without introducing noise from a non-numeric feature. Dummy coding is a standard term in statistics for this kind of encoding, emphasizing the reference category against which others are compared. A related idea is one-hot encoding, which also creates binary features for categories, but in many contexts dummy coding specifically highlights the reference-based representation used to avoid multicollinearity. Normalization or binarization, by contrast, refer to different data transformations that don’t capture the categorical-to-binary encoding concept.