roboticcam / machine-learning-notes Goto Github PK
View Code? Open in Web Editor NEWMy continuously updated Machine Learning, Probabilistic Models and Deep Learning notes and demos (2000+ slides) 我不间断更新的机器学习,概率模型和深度学习的讲义(2000+页)和视频链接
My continuously updated Machine Learning, Probabilistic Models and Deep Learning notes and demos (2000+ slides) 我不间断更新的机器学习,概率模型和深度学习的讲义(2000+页)和视频链接
请问dynamic_model.pdf的第33页公式推导最后一步是漏掉了p(y1|q1)吗
pdf第九页将ELOB改写成KL散度时应该是qi(Zi)和那个伪分布p(X,Z)之间的KL散度吧,所以之后的迭代公式里更新的是ln(qj(zj))而不是qj(zj)
首先感谢徐老师的视频和PPT,逻辑清晰,容易理解,帮助我省了很多的时间。
在学习《Variational Inference》时,根据我的理解发现ppt 的几处笔误,具体如下:
若以上4处“笔误” 有不对的地方,还请帮助指正
老师您好,麻烦您更新一下二维码,谢谢老师
推荐怎么提高学习效率的方法:方法
徐老师,2023年的网络课还在继续吗?还有的话麻烦发一下更新的二维码。
hello,课件有先后观看顺序吗?
想问一下 老师在课程里面用到的matlab代码 在哪里可以获得
周六直播二维码失效了,求更新
老师,您可以更新一下二维码吗,非常期待听您的课 , 谢谢!
失效了大佬
徐老师,您好!请问在哪里可以看到DeeCamp2018视频,特别是关于NCE和re-parameterization的内容。我在youtube,bilibili和优酷上都只看到变分推断的,非常感谢!
Greetings!
In the last 2 lines of the last slide in dynamic model, what do v_l and δ_(v, l) indicate?
Thanks!
徐老师可不可以出个hdp-hmm的视频讲解呀,还有关于hmm的扩展hsmm的讲解,公式推导之类的
Hi, Dr Xu, the QR code you posted is expired, can you please upload a new version?
Prof Xu said that he already published his code of Paper Geometrically-constrained balloon fitting for multiple
connected ellipses, and I'd like to debug it, but how can I get the code?
如题。期待徐老师讲解高斯过程~
You video help me a lot, and I really appreciate it.
先贴一下课程视频徐亦达机器学习:Kalman Filter 卡尔曼滤波,大概到7分13秒左右在推导$x_{t}|y_{1},...,y_{t-1}$时,使用了$P(x_{t}|x_{t-1})=N(Hx+B,Q)$这个公式,但是$x_{t}|y_{1},...,y_{t-1}$是基于条件$y_{1},...,y_{t-1}$而不是基于条件$x_{t},y_{1},...,y_{t-1}$因此不应该使用$P(x_{t}|x_{t-1})=N(Hx+B,Q)$这个公式直接求$x_{t}|y_{1},...,y_{t-1}$,而应该使用视频右上角的边缘积分去求
在求ln(q_j^* (Z_j))时,
ln(q_j^* (Z_j))=E_(i≠j) [ln(p(X,Z))]
在右侧计算时为什么只考虑与Zj有关的项,而将其他项直接去掉?例如,对于高斯函数,推导了:
ln(q_μ^* (μ))=E_(q_τ ) [ln(p(μ,τ│D)) ]
=-(E_(q_τ ) [τ])/2 [∑_(i=1)^n▒〖(x_i-μ)^2+λ_0 (μ-μ_0 )〗^2 ]+const
这个常数项其实是关于τ的函数,即可以写成:
ln(q_μ^* (μ))=-(E_(q_τ ) [τ])/2 [∑_(i=1)^n▒〖(x_i-μ)^2+λ_0 (μ-μ_0 )〗^2 ]+f(τ)
我觉得这个常数项不能直接忽略,因为:
ln(q_τ^* (τ))=E_(q_μ ) [ln(p(μ,τ│D)) ]
在求解ln(q_τ^* (τ))时用到q_μ^* (μ),则q_μ^* (μ)中含有τ的项不能忽略,这样理解有什么问题吗。
EM算法ppt第11页,对EM收敛 证明中的 Q函数以及H函数的推导有点疑惑 个人感觉 ln[p(x|theta)] 应该等于 ln[p(z,x|theta)]-ln[p(z|x,theta)]
徐老师你好,决策树 notes 文章里面table后面的那个数字显示不出来,都是问号??
rt, 有录播能不能分享一些链接
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.