形質遺伝を重視した突然変異の提案とその有効性

概要

論文の詳細を見る
The author has conducted research mainly into the use of genetic algorithm as a problem solving method applied to scheduling and combinational problem. A main feature of both kinds of problems is that their research objects do not permit multiple selection, therefore genetic manipulation has been carried out using order expression. However, when "crossover" and "mutation" are carried out in the order expression this breaks the good gene sequence from the previous generation. Therefore crossover" and "mutation" do not necessarily produce a very good individual. In another paper a different method called the "sub tour crossover" method was proposed. This method makes it possible to preserve the genotype as the genetic expression, therefore allowing the good genes from the previous generation to be kept. I wanted to make a mutation which emphasized characteristic heredity and devised and experimented with a new "swap type mutation". I tested this new genetic manipulation method and announced it at the JCKBSE'2000,but there was no proof of sufficient effectiveness at that time. However, after systematic experimentation with this method, the effectiveness of this technique was repeatedly confirmed and will be reported.
新潟国際情報大学の論文
2001-03-19