Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

生成对抗网络——原理解释和数学推导 - 黄钢的博客|ATuk Blog #33

Open
AlbertHG opened this issue May 6, 2018 · 13 comments

Comments

@AlbertHG
Copy link
Owner

AlbertHG commented May 6, 2018

https://alberthg.github.io/2018/05/05/introduction-gan/

Hello World

@NieYi
Copy link

NieYi commented Jun 2, 2019

(13)这个公式后面一个Pdata(x)少了一个帽子

@AlbertHG
Copy link
Owner Author

AlbertHG commented Jun 3, 2019

@NieYi
(13)这个公式后面一个Pdata(x)少了一个帽子

感谢纠错

@BZbyr
Copy link

BZbyr commented Dec 12, 2019

35-->36 怎么变成 log1/2的呀,是因为PG(x)那些是给定的无法积分吗?

@BZbyr
Copy link

BZbyr commented Dec 12, 2019

41-42 应该改成这样的吧

$$ \begin{align} G^{* } & =arg\ \underset{G}{min}\ \underset{D}{max}\ V(G,D) \ & =arg\ \underset{G}{min}\ (-2 log 2 + 2 JSD \left ( P_{data}(x) || P_G(x) \right)) \end{align} $$

@BZbyr
Copy link

BZbyr commented Dec 13, 2019

很详细!感谢!

@JiaqiuZhou
Copy link

收获很多,感谢

@Gi-gigi
Copy link

Gi-gigi commented Nov 22, 2020

图片看不到。。。

@Gi-gigi
Copy link

Gi-gigi commented Nov 24, 2020

说实话 没有图片 难以去想象

@Gi-gigi
Copy link

Gi-gigi commented Nov 24, 2020

请问 为何 Pdate+Pg=1?

@YihuaZhi
Copy link

YihuaZhi commented Dec 8, 2020

请问(17)式是怎么来的呀~怎么突然有个指示函数

@YihuaZhi
Copy link

YihuaZhi commented Dec 8, 2020

请问JS和JSD有什么区别,JS散度的取值范围不是0-log2吗,您写的是0-1?

@YihuaZhi
Copy link

YihuaZhi commented Dec 8, 2020

在《实际情况下的算法总结》中 更新D的时候,最大化应该是“+”?,如果是最小化交叉熵,应该是“-”?

@xuyanging
Copy link

有个问题,就是如果抛开GAN的结构看 只保留生成器,输入z,输出G(z)直接和真实数据x做比较,比如可以用交叉熵或者KL(G(z),x),这样为什么不行呢,哪里出现了问题呢

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

7 participants