Suppose that we are trying to build a decision tree to predict the responding behavior of customers, and want to figure out which attribute (Gender, Education, or Satisfaction) to be used first to split our data. To do so, we calculated the entropy of the subsets after splitting the original traing data set(40 data points) over each attribute and the results are shown below. The two numbers in the parathesis mean the entropy value and the number of data points in that subset. Based on this information, which attribute is the most informative, in terms of information gain, about the responding behavior of customers?

Gender: Male (0.32, 23), Femail (0.67, 17)

Eduction: Low (0.58, 16), Medium (0.87, 12), High (0.24, 12)

Satisfaction: Low (0.78, 6), Medium (0.69, 15), High (0.36, 19)

Group of answer choices

a. Satisfaction

b. They are equally informative

c. Education

d. Gender