Based on Entropy, which attribute (Temp or Windy) should be put to the root node (first level) of the decision tree?

Temp	Windy	Class Label (play)
95	false	No
90	true	No
77	false	Yes
70	true	Yes
68	false	Yes
60	true	No
55	true	No
50	false	Yes

Entropy= i-pi* log2 pi

a) Compute the overall entropy of the whole dataset without splitting (hint: check the number of instances in each class).

b) Compute the resulted entropy if we split on Windy.

c) If we split on Temp, what are the split points we should consider according to what we learned in class?

d) Compute the minimum entropy if we split on Temp.