I have a 4 dimensional feature and each of them are independent normal distributions. I want to calculate the bayesian error associated with this classifier. The covariance matrix and the mean have been given for multivariate normal distribution. I understand that I need to compute the error integral for the 4 dimensional space. However When I tried to compute the integral in mathematica, it throws up an error. I was wondering if there are any approximations or a workaround to compute the integral given all the above data?
Asked
Active
Viewed 487 times
3