From 926d0b37a01cf5e54deca8065bf9bc411c4850ef Mon Sep 17 00:00:00 2001 From: Rickustc <1072105344@qq.com> Date: Sat, 27 Apr 2024 19:57:44 +0800 Subject: [PATCH] Site updated: 2024-04-27 19:57:44 --- .../index.html" | 5 ++++- 2024/GAN/index.html | 5 ++++- .../index.html" | 5 ++++- .../index.html" | 5 ++++- archives/2023/09/index.html | 5 ++++- archives/2023/index.html | 5 ++++- archives/2024/04/index.html | 5 ++++- archives/2024/index.html | 5 ++++- archives/index.html | 5 ++++- categories/index.html | 7 +++++-- content.json | 2 +- curriculum/index.html | 5 ++++- index.html | 5 ++++- rick/cv.html | 5 ++++- rick/index.html | 5 ++++- rick/temp.html | 5 ++++- tags/DGM/index.html | 5 ++++- tags/LVM/index.html | 5 ++++- tags/VAE/index.html | 5 ++++- tags/index.html | 7 +++++-- .../index.html" | 5 ++++- "tags/\346\255\243\345\210\231\345\214\226/index.html" | 5 ++++- 22 files changed, 87 insertions(+), 24 deletions(-) diff --git "a/2023/\346\267\261\345\272\246\345\255\246\344\271\240\344\270\255\347\232\204\346\255\243\345\210\231\345\214\226/index.html" "b/2023/\346\267\261\345\272\246\345\255\246\344\271\240\344\270\255\347\232\204\346\255\243\345\210\231\345\214\226/index.html" index c855a81..b78f8f4 100644 --- "a/2023/\346\267\261\345\272\246\345\255\246\344\271\240\344\270\255\347\232\204\346\255\243\345\210\231\345\214\226/index.html" +++ "b/2023/\346\267\261\345\272\246\345\255\246\344\271\240\344\270\255\347\232\204\346\255\243\345\210\231\345\214\226/index.html" @@ -84,7 +84,10 @@

Reference

s.src = '//' + 'nothing' + '.disqus.com/embed.js'; s.setAttribute('data-timestamp', +new Date()); (d.head || d.body).appendChild(s); - })();
-

September 2023

-
-
-
-
+Categories - Rick Universe @@ -34,7 +34,10 @@ window.addEventListener('hashchange', switchTab, false); })(); -
-
-
-
+Tags - Rick Universe @@ -34,7 +34,10 @@ window.addEventListener('hashchange', switchTab, false); })(); -

贝叶斯的两个理解角度

When you have eliminated the impossible, whatever remains, however -improbable, must be the truth...

Read more

深度学习中的正则化

最近在看集成学习,集成学习中对基学习器的要求是和而不同,及满足两个条件:误差 \(\epsilon -<0.5\)且具有多样性,由此引出一系列多样性方法,其和正则化非常类似。借助课上提到的框架对正则化进行梳理。并且试图统一绝大部分ml和dl的正则化方法

Read more