Use of RBM for Identifying Linkage Structures of Genetic Algorithms
Use of RBM for Identifying Linkage Structures of Genetic Algorithms
复制标题
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
H. Handa
中科院分区：
文献类型：
--
作者：
H. Handa
连锁识别是遗传算法研究中一个非常重要的问题。如果发现了连锁结构，就可以改进编码方法，进行遗传操作。本研究采用受限玻尔兹曼机（RBM）来捕捉连锁结构。RBM的学习数据包括通过参考通过非线性检查的联动识别的非线性检查而生成的数据。将RBM中隐藏层的神经元的激活值和相应的学习数据可视化。我们可以很容易地找出链接结构，通过使用所产生的图像。
—Linkage Identification is a quite important in Genetic Algorithm studies. If we found linkage structures, we can improve coding methods, genetic operations. The Restricted Boltzmann Machine (RBM) is adopted to capture linkage structures in this study. The learning data of the RBM consist of data generated by referring to the nonlinearity check of the Linkage Identification by non-Linearity Check. The activation values of the neurons in the hidden layer in the RBM and corresponding learning data are visualized. We can easily find out linkage structures by using the resultant images.