An example of what type of target variable is purchase or no-purchase?

The target variable in question, which reflects whether a purchase is made or not (i.e., purchase or no-purchase), is characterized as binary. This is because there are only two distinct categories or classes within the variable—either the event occurs (purchase) or it does not occur (no-purchase).

In the context of classification tasks in predictive modeling, binary variables are often the focus when the outcome can only fall into two possible states. This structure allows for straightforward analysis using various statistical methods and algorithms designed for binary classification.

Moreover, when considering the options of interval, ordinal, and nominal types, binary is the best fit since interval refers to continuous data with a defined scale, ordinal implies a ranking where order matters but intervals between ranks are not uniform, and nominal pertains to categories without an inherent order. Therefore, identifying the target variable as binary reflects its core nature of having two mutually exclusive outcomes, aligning perfectly with its conceptual understanding in statistical analysis and machine learning.