- 62
- 117 159
MLSS Africa
เข้าร่วมเมื่อ 27 มี.ค. 2018
Machine Learning Summer School
MLSS 2019 David Blei: Variational Inference: Foundations and Innovations (Part 1)
David Blei
Topic: Variational Inference: Foundations and Innovations (Part 1)
Topic: Variational Inference: Foundations and Innovations (Part 1)
มุมมอง: 14 065
วีดีโอ
MLSS 2019 OVERVIEW
มุมมอง 5765 ปีที่แล้ว
An overview of the MLSS 2019 held in Stellenbosch, South Africa.
How to write a great research paper Part 2
มุมมอง 1.3K6 ปีที่แล้ว
MLSS 2019. Stellenbosch, South Africa. How to write a great research paper (part 2)
SAFA MOHAMED - "My MLSS 2019 experience"
มุมมอง 1506 ปีที่แล้ว
MLSS 2019, held at Stellenbosch, South Africa.
Ben Herbst Mcelroy Hoffman Data Science in Practice
มุมมอง 3056 ปีที่แล้ว
Ben Herbst Mcelroy Hoffman Data Science in Practice
Michelle Lochner Big Data Challenges for Next Generation Astronomy
มุมมอง 1986 ปีที่แล้ว
Michelle Lochner Big Data Challenges for Next Generation Astronomy
RAIMA CAROL APPAW - "MY MLSS 2019 experience"
มุมมอง 1266 ปีที่แล้ว
MLSS 2019 hosted in Stellenbosch, South Africa
AYA SALAMA - "My MLSS 2019 experience"
มุมมอง 2546 ปีที่แล้ว
MLSS 2019 hosted in Stellenbosch, South Africa
SAFA MOHAMED - "My MLSS 2019 experience"
มุมมอง 866 ปีที่แล้ว
MLSS 2019 hosted in Stellenbosch, South Africa
Ruslan Salakhutdinov Introduction to Deep Learning
มุมมอง 6956 ปีที่แล้ว
Ruslan Salakhutdinov Introduction to Deep Learning
Bettina Berendt What’s the hype about ethical AI?
มุมมอง 1206 ปีที่แล้ว
Bettina Berendt What’s the hype about ethical AI?
Stefan van der Walt How to write good scientific software
มุมมอง 1906 ปีที่แล้ว
Stefan van der Walt How to write good scientific software
Dimitrios Lyras Knowledge discovery via AI and creative thinking
มุมมอง 746 ปีที่แล้ว
Dimitrios Lyras Knowledge discovery via AI and creative thinking
Iain Murray Probabilistic modeling and inference meets Deep Learning Part 2
มุมมอง 5636 ปีที่แล้ว
Iain Murray Probabilistic modeling and inference meets Deep Learning Part 2
Sharon Goldwater Speech and language processing for low resource languages
มุมมอง 7026 ปีที่แล้ว
Sharon Goldwater Speech and language processing for low resource languages
Bernhard Schölkopf Causal Discovery Part 1
มุมมอง 2.9K6 ปีที่แล้ว
Bernhard Schölkopf Causal Discovery Part 1
Bernhard Schölkopf Causal Discovery Part 2
มุมมอง 1.4K6 ปีที่แล้ว
Bernhard Schölkopf Causal Discovery Part 2
Shakir Mohamed Planting the Seeds of Probabilistic Thinking Part 2
มุมมอง 9086 ปีที่แล้ว
Shakir Mohamed Planting the Seeds of Probabilistic Thinking Part 2
Zoubin Ghahramani AMA Ask Me Anything Skype
มุมมอง 3766 ปีที่แล้ว
Zoubin Ghahramani AMA Ask Me Anything Skype
Shakir Mohamed Planting the Seeds of Probabilistic Thinking Part 3
มุมมอง 6376 ปีที่แล้ว
Shakir Mohamed Planting the Seeds of Probabilistic Thinking Part 3
David Blei Variational Inference Foundations and Innovations Part 2
มุมมอง 4.5K6 ปีที่แล้ว
David Blei Variational Inference Foundations and Innovations Part 2
Luc De Raedt Statistical Relational Learning and Probabilistic Logic Programming Part 2
มุมมอง 6686 ปีที่แล้ว
Luc De Raedt Statistical Relational Learning and Probabilistic Logic Programming Part 2
Please, delete these videos. They are useless. No Quality for great lectures.
th-cam.com/video/ZsH2zc71t78/w-d-xo.html
This is so good ! Wonderful lecture!
Video is too focused on just the speaker. I need to see what he is pointing at
how to get ppt
Da bes
42:30 x_i is dependent on z_j. But x_i is conditionally independent from z_j if beta is known.
Such a nice explanation. And how good the slides were!
The main things I get from here: • conditioning: 1:08:15 • (didn't find it anywhere else) the idea about underfitting/overfitting with GP regression: 1:22:05 - 1:24:45
can someone help me get the slides?
Varun, try www.cs.columbia.edu/~blei/talks/Blei_VI_tutorial.pdf
Best math video I've seen in a while.
@MLSS Africa. Are the slides availble for this? Video quality is bad
Can we take a moment to appreciate Blei's dry sense of humor
Bruh makes me sad hindi is low resource language maybe we need some billionaire to help create resources
Why keep a horrible video for such a great presentation ?
Exactly.
Spectacular presentation!
Grandmaster Latex skills
The video quality for this is quite a bit worse than for the first 2 parts!
It gets better around 31:00
this is Blei, not Blie
shades being thrown 34:50
Fantastic
Wow, this is an extremely good explanation in so many ways. I love the GAN explanation! Thank you!!
42:10 When Shakir goes beast mode...
Hi @MLSS Africa, Are these slides available anywhere?
drive.google.com/file/d/1RNrgDs5xw-9HTjikFU1L0iO1PBMDaGwE/view
www.gatsby.ucl.ac.uk/~gretton/papers/cardiff.pdf
Great talk!
best explanation on OT! thanks a lot
At 1:07:06, I thought the BDD is encoding the disjunctions on the left, instead he shows the BDD encodes the formulas on the right ?
Such a valuable talk. It would be great if you could share the slides.
Found one on their website! drive.google.com/drive/folders/1mHqPevddlosT7TnqumDBq4e-fIw8sot_
@@Rowing-li6jt Thank you!
Where are the slides? The ones on the website don't include Neil's talks?
Damn it, this should have a second part as well :(
What happened to this video, why there are blanks?
Are the slides posted somewhere?
Slides are available here: inverseprobability.com/talks/slides/2019-01-09-gaussian-processes.slides.html and more details are also on Neil's website: inverseprobability.com/talks/notes/gaussian-processes.html
Slides anywhere?
drive.google.com/drive/folders/1mHqPevddlosT7TnqumDBq4e-fIw8sot_
www.shakirm.com/slides/MLSS2018-Madrid-ProbThinking.pdf
Thanks for uploading the video. A great and interesting talk.
Only one like? Shame cause to me it was illuminating ;)
His laser pointer wasn't illuminating much, as far as I could observe.
Part 1, please?
26:50 Good question about "extending causal structures/modelling to more than just directed graphs". I'm surprised that the panelists didn't mention probabilistic programming. Here is a motivation for why you need to move beyond the formalism of graphs/PGMs: pinouchon.github.io/ai/2016/05/20/thick-arrows.html
@Crouzier Benjamin, extending causal models beyond the realm of DAG's defeats the purpose of causality and studying causal inference, what Josh has been doing is trying to capture that same underlying causal structure with probabilistic programming in the hope that these probabilistic approximations would at least give good approximate answers about the hidden causal structure which actually accounts for how the data is being generated because there are two things to keep in mind once you know the true causal structure you may not still know about all the underlying random variables that may be in play, if you know all the variables you might then not know the actual underlying "structure", so it is essentially a fight between being complete and consistent and probabilistic programming is a small step in that direction, and there are other approaches to such as using variational Bayesian inference built on top pf mean filed approximations to capture the same things.
oh wow what a great panel
The slide on the log normalizer is missing the base density h(x).
1 dislike from frequentist sect
Why slides are not available for this talk?
github.com/nowozin/mlss2018-madrid-gan
where can I learn more about Probabilistic programming?
search problog in google.
We need some problems in the OT book. Can't get enough of this stuff!
Some comments Bayesian foundations I made here: livelogic.blogspot.com/2019/05/julia-galef-on-aumanns-agreement-theorem.html
Some comments I made on philosophers programming 😂 and on resonance etc here: livelogic.blogspot.com/2019/05/john-skilling-on-probability-and.html?m=0
Some comments I made on exploration algorithms here: livelogic.blogspot.com/2019/05/john-skilling-on-big-spaces.html
They need to fire the intern controlling the slides
It seems that the person controlling the slides had to install Chrome during the presentation because initially the equations were not showing in Safari.
And the one moving the video camera too
13:45 He uses the wrong notation. Bit cross in mathematics means *cross product*
Hello MLSS, where is part 1 of this talk, the panel discussion?