A simple example for continuous inputs
• Assume that the input vectors for each class are from a
Gaussian distribution, and all classes have the same
covariance matrix.
• For two classes, C1 and C0, the posterior is a logistic: