CAJ | 학술논문

提出一种改进的蚁群算法（即交叉验证法与蚁群算法的有效结合）对支持向量机的参数进行优化。首先通过交叉验证法得到支持向量机参数，然后在这些参数中确定蚁群算法的搜索空间，最后在该区间上选取最优参数。仿真结果表明：使用该方法优化支持向量机参数可以在较短时间内寻找到最优解，进而缩短搜索时间，提高识别准确率。
제출일충개진적의군산법（즉교차험증법여의군산법적유효결합）대지지향량궤적삼수진행우화。수선통과교차험증법득도지지향량궤삼수，연후재저사삼수중학정의군산법적수색공간，최후재해구간상선취최우삼수。방진결과표명：사용해방법우화지지향량궤삼수가이재교단시간내심조도최우해，진이축단수색시간，제고식별준학솔。
An improved ant colony algorithm ( combination of the cross validation method and the ant colony algorithm) was proposed to optimize SVM ( support vector machine) parameters.First of all,it makes use of the cross validation method to obtain SVM parameters,and then it determines ant colony algorithm search space in these parameters and selects optimal parameters at this interval.The simulation results show that this method can optimize SVM parameters and find optimal solution in a short period of time, and can reduce the search time as well as improve the accuracy rate of recognition.