Research and Implementation of Text Classification Model Based on Combination of DAE and DBN
Zhenyu Yang, Xue Pang · 2017
This paper presents a text classification method based on Denoising AutoEncoders(DAE) and Deep Belief Nets(DBN) for the deep learning model. Firstly, using the Denoising AutoEncoders(DAE) to learn other forms of expression of the initial feature of the text. And then use the Deep Belief Nets(DBN) to project the data and reduce the dimension. So that the text data can be expressed more further and better. Finally, the Softmax regression layer was used for classification. The experimental results show that the extracted text feature is applied to the text classification, which significantly improves the classification effect.