In the discriminant analysis method, what can be the data type of the attribute that is to be predicted?

• A. Numeric
• B. Categorical
• C. Text
• D. All of the above

Answer: (D)

Discriminant analysis is used for classification, so the thing you are predicting is a class label, i.e., a categorical variable. In practice, the attributes you use as predictors can be numeric or categorical (often after encoding), and even text data can be converted into numeric features before applying the method. The final predicted outcome, however, is a category. So while the target itself is categorical, the data you work with in the analysis pipeline can include numeric, categorical, and text data once it’s transformed for modeling, which is why the option All of the above is considered in broader practice.