The purpose of this paper is to propose a method to do classification by linear discriminant analysis using genetic algorithms. The author has provided fairly detailed experimental results of the execution of the algorithm on real-world data. He used some variation of the GENESIS package to simulate the search. The paper is fairly well written, but it contains nothing conceptually new, and the claims made in the paper about the superiority of the genetic algorithm approach as well as the robustness of the approach are at best not fully supported by evidence (theoretical or experimental).