Hierarchical Attention Networks for Document Classification阅读笔记
2019-04-09 23:43
791 查看
文章目录
模型结构
Hierarchical Attention
encoder采用的双向GRU
Word Encoder
Word Attention
Sentence Encoder
Sentence Attention
Document Classification
实验
数据集
- Yelp reviews
- IMDB reviews
- Yahoo answers
- Amazon reviews
参数
首先, 把文档切分成句子, 并用CoreNLP分词.
- train: val: test = 80%: 10%: 10%
- word_embedding: 200
- GRU dimension: 50(双向之后拼接成100)
- batch_size: 64
- optimizer: SGD with momentum(0.9)
实验结果
相关文章推荐
- Hierarchical Attention Network for Document Classification阅读笔记
- 论文笔记《Hierachical Attention Networks for Document Classification》
- Implementation of Hierarchical Attention Networks for Document Classification的讲解与Tensorflow实现
- Recurrent Convolutional Neural Networks for Text Classification阅读笔记
- 【论文阅读笔记】MULTI-SCALE DENSE NETWORKS FOR RESOURCE EFFICIENT IMAGE CLASSIFICATION
- 阅读笔记(paper+code):Residual Attention Network for Image Classification
- 论文阅读笔记:Transformation Networks for Target-Oriented Sentiment Classification
- 阅读笔记(paper+code):Residual Attention Network for Image Classification
- A Committee of Neural Networks for Traffic Sign Classification 阅读笔记
- Hierarchical Attention Network for Document Classification--tensorflow实现篇
- 文献阅读笔记——Non-Associative Higher-Order Markov Networks for Point Cloud Classification
- FeUdal Networks for Hierarchical Reinforcement Learning 阅读笔记
- 【DRCN阅读】Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation. 论文笔记
- 论文阅读笔记:MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
- 论文阅读笔记:Fully Convolutional Networks for Semantic Segmentation
- 深度学习笔记(一)空间金字塔池化阅读笔记Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
- 论文阅读笔记:Fully Convolutional Networks for Semantic Segmentation
- LSTM Recurrent Neural Networks for Short Text and Sentiment Classication文章阅读笔记
- 【论文阅读笔记】CVPR2015-Long-term Recurrent Convolutional Networks for Visual Recognition and Description
- 论文阅读笔记 | (ICCV 2017) Multi-Attention CNN for Fine-Grained Image Recognition:MA-CNN