TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Stanford Stats 385: Theories of Deep Learning

217 点作者 capocannoniere超过 7 年前

6 条评论

inputcoffee超过 7 年前
I feel like asking: did they solve the problem?<p>Let me see if I can state the problem: Neural Networks are non-linear because of their activation functions. You need a differentiable function in order to take the derivative so you can back-prop the error, more or less.<p>The consequence of the non-linearity is that you can&#x27;t do some kind of short hand calculation to figure out what the network will do. You have to crank through the network to see the result. There is no economy of thought. That is to say, there is no theory.<p>I am excited that they are working on it, but I would love to have a summary or overview of how they approach what I consider to be the basic problem.
评论 #15647090 未加载
评论 #15647151 未加载
评论 #15647060 未加载
评论 #15646620 未加载
eduren超过 7 年前
Does anybody know if they plan on releasing the lecture videos? I couldn&#x27;t find them on the site and this looks very interesting.
评论 #15649849 未加载
AlexCoventry超过 7 年前
I&#x27;m a bit surprised that Soatto&#x27;s and Tishby&#x27;s papers aren&#x27;t on the reading list (<a href="https:&#x2F;&#x2F;stats385.github.io&#x2F;readings" rel="nofollow">https:&#x2F;&#x2F;stats385.github.io&#x2F;readings</a>). I think they have some of the most interesting theories, at the moment, about why Deep Learning works.<p><a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1706.01350" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1706.01350</a><p><a href="https:&#x2F;&#x2F;openreview.net&#x2F;pdf?id=ry_WPG-A-" rel="nofollow">https:&#x2F;&#x2F;openreview.net&#x2F;pdf?id=ry_WPG-A-</a>
评论 #15653592 未加载
kleiba超过 7 年前
What a privilege to study at that college.
评论 #15648151 未加载
评论 #15650381 未加载
tequila_shot超过 7 年前
Can general public attend these sessions?
评论 #15646478 未加载
pwaivers超过 7 年前
The pictures on the first page look exactly like what I expect graduate students and professors of Stats to look like.