Question 9
Domain 2: Explore data and run experimentsYou are using Azure Machine Learning Studio to perform feature engineering on a dataset. You need to normalize values by converting a numeric feature into a column of discrete bins. Does applying Entropy Minimum Description Length (MDL) binning meet this goal?
Correct answer: B
Explanation
Entropy MDL binning is a supervised discretization method that “uses class information to determine bin boundaries,” so it creates bins from labeled data rather than normalizing values. Normalization means scaling or standardizing numeric values, not converting them into discrete intervals.
Why each option is right or wrong
A. Yes, because Entropy MDL binning normalizes numeric values into a grouped bin representation.
B. No, because Entropy MDL binning is used for supervised discretization, not for value normalization in the sense of scaling or standardizing features.
Entropy MDL binning in Azure Machine Learning Studio is a supervised discretization method: it uses the class label to choose cut points that minimize description length, so it is appropriate when you want to turn a continuous variable into intervals based on target information. The Azure ML discretization operator does not perform numeric normalization such as min-max scaling or z-score standardization, which are the usual ways to normalize values; therefore it does not satisfy the requirement to normalize a feature into bins.