Genetic-CNN-master

所属分类人工智能/神经网络/深度学习
开发工具:Python
文件大小:110KB
下载次数:2
上传日期:2020-04-30 10:09:41
上 传 者崔宁敏
说明:  为了自动学习深度网络结构,随着网络结构的数量随网络中层数的增加而呈指数增长,这启发我们采用遗传算法来有效地遍历此较大的搜索空间。我们首先提出一种编码方法,以固定长度的二进制字符串表示每个网络结构,然后通过生成一组随机个体来初始化遗传算法。在每一代中,我们定义标准的遗传操作,例如选择,突变和交叉,以消除弱势个体,然后产生更具竞争力的个体。
(In order to learn the deep network structure automatically, the number of network structures increases exponentially with the increase of the number of middle layer, which inspires us to use genetic algorithm to traverse this large search space effectively.We first propose a coding method to represent each network structure as a binary string of fixed length, and then initialize the genetic algorithm by generating a set of random individuals.In each generation, we define standard genetic operations, such as selection, mutation and crossover, to eliminate vulnerable individuals and then produce more competitive ones.)

文件列表:[举报垃圾]
Genetic-CNN-master, 0 , 2019-07-20
Genetic-CNN-master\Genetic CNN.ipynb, 13397 , 2019-07-20
Genetic-CNN-master\GeneticCNN.py, 9906 , 2019-07-20
Genetic-CNN-master\LICENSE, 11357 , 2019-07-20
Genetic-CNN-master\README.md, 1563 , 2019-07-20
Genetic-CNN-master\dag.py, 6715 , 2019-07-20
Genetic-CNN-master\ga-cnn.png, 103250 , 2019-07-20

近期下载者

相关文件

评论我要评论

收藏者