A RATIO MODEL FOR DISCRIMINANT-ANALYSIS USING LINEAR-PROGRAMMING

The innovative hybrid model for discriminant analysis via linear progr amming, was introduced along with the beta normalization, which was su bsequently replaced by a new normalization. The primary weakness of th e beta normalization, and the reason for replacing it with the new nor malization, is that it distorts solutions and consequently does not al ways find the best solution. It is shown here, that unfortunately, bot h normalizations are affected by distance distortion. In addition, whe ther a model finds the best solution is highly dependent on the criter ion or criteria by which ''best solution'' is defined. A ratio criteri a and associated ratio model are proposed, which avoid the problems as sociated with distance distortion. The nonlinear ratio may be lineariz ed in much the same way that Data Envelopment Analysis is linearized.