Use of Linear Programming Approaches For the Classification Problem in Discriminant Analysis

The classification problem in discriminate analysis involves the construction a rule for the assignment of an observation into one of several groups. Various discriminate analysis techniques have been proposed in the fields of statistics, management science and artificial intelligence. This paper introduces two most widely used linear programming (LP) formulations to the classification problem. The results of experimental comparisons of LP and the statistical approaches are reported.

Keywords:

Linear programming , symmetric trapezoidal fuzzy numbers , fuzzy simplex methods