Classification Error
The classification error is the overall error we make with a given model for a given dataset. It is calculated using the formula:
$$ p(error) = \sum_{i=1}^C p(error|\omega_i)p(\omega_i) $$ This can better be visualized by integrating the area under the distribution of two classes that are to opposing sides of the decision boundary:
