Improving deep convolutional neural networks with mixed maxout units

Show other versions (1)

The maxout units have the problem of not delivering non-max features， resulting in the insufficient of pooling operation over a subspace that is composed of several linear feature mappings,when they are applied in deep convolutional neural networks.The mixed maxout (mixout) units were proposed to de...

Full description

Bibliographic Details
Published in:Tongxin xuebao
Main Authors: Hui-zhen ZHAO, Fu-xian LIU, Long-yue LI, Chang LUO
Format: Article
Language:Chinese
Published: Editorial Department of Journal on Communications 2017-07-01
Subjects:
deep learning;convolutional neural network;maxout units;activation function
Online Access:http://www.joconline.com.cn/thesisDetails#10.11959/j.issn.1000-436x.2017145

Internet

http://www.joconline.com.cn/thesisDetails#10.11959/j.issn.1000-436x.2017145

Similar Items