The comparison of the average ARI and %CA scores produced by 50 consecutive runs of MOGASVM with ensemble of kernel functions and MOGASVM with individual kernel functions for all the datasets.

The comparison of the average ARI and %CA scores produced by 50 consecutive runs of MOGASVM with ensemble of kernel functions and MOGASVM with individual kernel functions for all the datasets.针对所有数据集，对比了采用核函数集成的多目标遗传算法支持向量机（Multi-Objective Genetic Algorithm Support Vector Machine，MOGASVM）与采用单一核函数的MOGASVM连续50次运行所得到的平均调整兰德指数（Adjusted Rand Index，ARI）及分类准确率百分比（%CA）得分。