Assume that in a binary classification problem with one feature X, the distribution of X in...

Transcribed Image Text:

Assume that in a binary classification problem with one feature X, the distribution of X in class k = 1 is fi(x) exp x > 0 20? and the distribution of X in class k = 2 is -(In x – µ2)² 203 1 f2(x) exp x/2702 x > 0 (a) Are there any conditions under which the discriminant function is a linear function of x? (b) If o1 = 02 = 1, µ1 = 10, and 71 = T2 = 0.5, in what class will x = 10 be classified? Assume that in a binary classification problem with one feature X, the distribution of X in class k = 1 is fi(x) exp x > 0 20? and the distribution of X in class k = 2 is -(In x – µ2)² 203 1 f2(x) exp x/2702 x > 0 (a) Are there any conditions under which the discriminant function is a linear function of x? (b) If o1 = 02 = 1, µ1 = 10, and 71 = T2 = 0.5, in what class will x = 10 be classified?