Consider the following source of (attribute-value, class) pairs.
S = {(20, C1), (21, C1), (22, C1), (24, C1), (24, C2), (25, C1),
(25, C2), (26, C1), (27, C1), (28, C1), (28, C2), (29, C1), (29,
C2), (30, C2), (33, C1), (34, C2), (35, C2), (39, C2), (41, C2),
(42, C2)}
(a) Find the source entropy. (b) Find the optimum threshold to
separate the two classes, and which maximizes the information
gain.
Using Matlab, please.