通信学报2017，Vol.38Issue(7)：105-114,10.DOI:10.11959/j.issn.1000-436x.2017145
基于混合maxout单元的卷积神经网络性能优化
Improving deep convolutional neural networks with mixed maxout units
摘要
Abstract
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 convolu-tional neural networks. The mixed maxout (mixout) units were proposed to deal with this constrain. Firstly, the exponen-tial probability of the feature mappings getting from different linear transformations was computed. Then, the averaging of a subspace of different feature mappings by the exponential probability was computed. Finally, the output was ran-domly sampled from the max feature and the mean value by the Bernoulli distribution, leading to the better utilizing of model averaging ability of dropout. The simple models and network in network models was built to evaluate the perfor-mance of mixout units. The results show that mixout units based models have better performance.关键词
深度学习/卷积神经网络/maxout单元/激活函数Key words
deep learning/convolutional neural network/maxout units/activation function分类
信息技术与安全科学引用本文复制引用
赵慧珍,刘付显,李龙跃,罗畅..基于混合maxout单元的卷积神经网络性能优化[J].通信学报,2017,38(7):105-114,10.基金项目
国家自然科学基金资助项目(No.61601499) The National Natural Science Foundation of China (No.61601499) （No.61601499）
