学习笔记 | NIPS 2021 regularization cocktail 调优的 MLPs 在表格数据上优于 GBDTs | Regularization is All Your Need

Posted 2023-01-02 叶庭云

文章目录

• 一、前言
• 二、Motivation
• 三、Regularizations
• 四、Methods
• 五、实验
• 六、总结

CSDN 叶庭云：https://yetingyun.blog.csdn.net/

---

一、前言

Well-tuned Simple Nets Excel on Tabular Datasets

Even simple neural networks can achieve competitive classification accuracies on tabular datasets when they are well regularized, using dataset-specific regularization cocktails found via standard hyperparameter optimiz