Re: Bayes decision boundary between 3 classes?



CORRECTED FOR THE UNFORGIVABLE SIN OF TOPPOSTING

Question wrote:
Greg heath wrote:
Consider the three equations

P(i)*p(x|i) = P(j)*p(x|j)

Should we have 3 decision boundaries for this example?

Yes.

P(i)*p(x|i) = max{ P(j)*p(x|j), P(k)*p(x|k) }

Hope this helps.

Greg

.