long-term engagement Optimize介绍

Reading time ~1 minute

<Returning is Believing: Optimizing Long-term User Engagement in Recommender Systems>

3.方法

我们提出了从顺序决策最优化的角度,在推荐中改进long-term user engagement。

参考

DTS介绍

youku在《Deep Time-Stream Framework for Click-Through Rate Prediction by TrackingInterest Evolution》提出了一个兴趣演进的框架。# 摘要CTR预测在像视频推荐等工业应用中是一个必要...… Continue reading

taobao Privileged Features Distillation介绍

Published on December 16, 2020

DFN介绍

Published on December 15, 2020