[其他] 计算视觉What Can ResNet Learn Efficiently, Going Beyond Kernels [推广有奖]

0关注
22粉丝

已卖：1945份资源

大师

20%

还不是VIP/贵宾

威望: 1 级
论坛币: 546 个
通用积分: 575.4231
学术水平: 66 点
热心指数: 130 点
信用等级: 37 点
经验: 113870 点
帖子: 6738
精华: 0
在线时间: 2932 小时
注册时间: 2022-8-18
最后登录: 2025-12-19

楼主

2023Hua

发表于 2025-8-11 11:38:20 |AI写论文

是否 +2 论坛币

k人参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

What Can ResNet Learn Efficiently,
         Going Beyond Kernels?

      Zeyuan Allen-Zhu             Yuanzhi Li
      Microsoft Research AI          Carnegie Mellon University
      zeyuan@csail.mit.edu          yuanzhil@andrew.cmu.edu

                     Abstract
   How can neural networks such as ResNet efficiently learn CIFAR-10 with test
   accuracy more than 96%, while other methods, especially kernel methods, fall
   relatively behind? Can we more provid ...