Deep Interest Network(DIN)

源码笔记：用Attention让每个用户预测关注的兴趣点（行为向量）不同。不过感觉这份源码可读性不太好。

建议阅读顺序：build_dataset-->input-->Dice-->model-->train

原paper：

@article{Zhou2017Deep,
  title={Deep Interest Network for Click-Through Rate Prediction},
  author={Zhou, Guorui and Song, Chengru and Zhu, Xiaoqiang and Ma, Xiao and Yan, Yanghui and Dai, Xingya and Zhu, Han and Jin, Junqi and Li, Han and Gai, Kun},
  year={2017},
}

原code： https://github.com/zhougr1993/DeepInterestNetwork