【KDD2020Tutorial】自动推荐系统，AutomatedRecommendationSystem

点击上方&＃xff0c;选择星标或置顶&＃xff0c;不定期资源大放送&＃xff01;

阅读大概需要5分钟

Follow小博主&＃xff0c;每天更新前沿干货

随着推荐任务的日益多样化和推荐模型的日益复杂&＃xff0c;开发出一套能够很好地适应新的推荐任务的推荐系统变得越来越具有挑战性。在本教程中&＃xff0c;我们将重点讨论自动机器学习(AutoML)技术如何有益于推荐系统的设计和使用。具体地说&＃xff0c;我们将从一个完整的范围开始描述什么是可以自动推荐系统。然后&＃xff0c;我们将在此范围内对特征工程、超参数优化/神经结构搜索和算法选择三个重要的主题进行详细阐述。将介绍、总结和讨论这些主题下的核心问题和最近的工作。

1. Introduction to AutoML
   

2. Phase I: Automated Model Design / Neural Architecture Search

3. 1. Efficient Neural Interaction Functions Search

   2. Automated Model Search for Collaborative Filtering
      

4. Phase II: Hyper-parameter Optimization for Recommendation

5. 1. Regularization Automatic: Framework and Method

   2. Embedding Size Automatic: Methods and Directions

   3. Learning Rate and Other Parameters Optimization
      

6. Phase III: Feature Engineering for Recommendation

7. 1. AutoCross: Automatic Feature Crossing for Tabular Data

   2. AutoFM: Automatic Feature Selection for FM
      

8. Phase IV: Automated Exploitation of Rich Side Information

9. 1. Automated Knowledge Graph for Recommendation

   2. Automated Graph Neural Networks for Recommendation
      

10. Phase V: Conclusion and open discussions

传送门

便捷下载&＃xff0c;请关注深度学习技术前沿公众号

• 后台回复“ARS” 就可以获取《【KDD2020-Tutorial】自动推荐系统&＃xff0c;Automated Recommendation System》下载链接