#1 show your work. If you write code to implement this, show your calculations and compare it with the code solution.

1 (40) Consider the training examples shown in Table 1 for a binary classification problem.

a1	a2	a3	Class
T	T	1	+
T	T	5	+
T	F	4	-
F	F	3	+
F	T	8	-
F	F	8	-
T	F	7	+
F	T	4	-
F	F	12	+
T	T	10	-

Table 1. Sample training data for a binary classification problem for question #1.

(a) What is the entropy (information content) of the entire training data?

(b) What are the information gains of a1 and a2 and report which one gives a better split.

(c) Compute Gini index for a1 and a2 and report which one gives a better split.

(d) Compute information gain for a3 for each potential split point and report the best split point for a3.