摘要
以深度学习中的自编码器为基础,堆积稀疏自编码形成深度网络,以贪心算法逐层训练实现特征层的层抽象映射,监督训练Softmax分类器,然后使用BP(反向传播)算法优化权值,构建SAE(栈式自编码器),并重点对其特征表达能力进行探究。第一,以单层自编码器作为特征表达的基础模块,探究自编码器中隐藏层节点数对特征表达能力影响;第二,重点探究对于多层自编码器的理解,堆积自编码器是否是一个好的获取理想特征表达的途径,主要评估多层自编码器对于特征表达的准确性和稳定性影响。基于MNIST(美国国家标准与技术研究所数据库)数据集的实验与PCA(主成分分析)、LLE(局部线性嵌入)、BP算法特征表达能力进行对比分析,验证栈式自编码器特征表达能力的有效性。
引文
1 SCHMIDHUBER J.Deep learning in neural networks:An overview[J].Neural Networks,2015(61)61:85-117.
2朱成,刘海强,朱峰,孙启新.电信大数据的数据挖掘关键技术分析与探讨[J].电信快报,2018(6):22-24.
3 WANG Y S,YAO H X,ZHAO S C.Auto-encoder based dimensionality reduction[J].Neurocomputing,2016,184:232-242.
4 LIAO B,XU J G,LV J T,ZHOU S L.An Image Retrieval Method for Binary Images Based on DBN and Softmax Classifier[J].IETE Technical Review,2014,32(4):292-303.
5 HINTON G E,OSTINDERO S,TEH Y W.A fast learning algorithm for deep belief nets[J].Neural Computation,2006,18(7):1527-1554.
6 YIN H P,JIAO X P,CHAI Y,et al.Scene classification based on single-layer SAE and SVM[J].Expert Systems With Applications,2015,42(7):3368-3380.
7 WANG S H,XIANG J W,ZHONG Y T,A data indicatorbased deep belief networks to detect multiple faults in axial piston pumps[J].Mechanical Systems and Signal Processing,2018(112):154-170.
8 VINCENT P,LAROCHELLE H,LAJOIE I,et al.Stacked Denoising Autoencoders:Learning Useful Representations in a Deep Network with a Local Denoising Criterion[J].Journal of Machine Learning Research,2010,11(12):3371-3408.