ABSTRACT
We aim at finding out the weighing vector that when applied to some newly observed and unclassified person will assign him to one or other of the classes with the smallest probability of error. The vector of weights (w) which provides the optimum assignment is given by the following relation:
where V is the weighted average of the variance-covariance matrix. It is given by the following relation:
1 2 ( 2) W W
n n +
+ −
where d is the vector of differences between the t-pairs of means of the two groups.
