- Consider a 2-dimensional, 2-class problem where the patterns are normally
distributed with the same covariance matrix, and
P(C1) = P(C2) = 0.5.
and the mean vectors are:
μ = [1;1]   μ = [2;1]
- Find the decision boundary between the two classes.
- Find the probability of an error.
- For the previous problem, assume that the classes have different
covariance matrices:
C1 = [2 1;1 4] C2 = [4 0.5;0.5 1]
and find the decision boundary between the two classes.
- For the previous problems, assume that both classes have the same
covariance matrix:
and the a priori probabilities P(C1) = 0.25, P(C2) = 0.75
and find the decision boundary.
- For the previous problem, assume the following loss values:
l11 = l (choose 1 | 1) = l22 = 0
l12 = 0.5   l21 = 2.5
and determine the decision boundary.