Bayesian or Frequentist, Which Are You? By Michael I. Jordan (Part 1 of 2)
HTML-код
- Опубликовано: 7 авг 2024
- Recorded: September 2009
at the Department of Electrical Engineering and Computer Sciences, UC Berkeley.
Part 1 of 2.
______________________________________
download the Slides:
videolectures.net/site/normal_...
0:00 Are You a Bayesian or a Frequentist?
1:28 Statistical Inference
6:37 Machine Learning
10:11 Decison-Theoretic Perspective -1
13:01 Decison-Theoretic Perspective -2
18:19 Decison-Theoretic Perspective -3
19:55 Coherence and Calibration
25:23 The Bayesian World
27:04 Subjective Bayes
34:03 Objective Bayes
36:44 Frequentist Perspective
39:21 Frequentist Activities
50:35 Outline -1
50:50 Surrogate Loss Functions
52:02 Motivating example: Decentralized Detection
53:18 Decentralized Detection
54:27 Decentralized Detection (cont.)
56:57 Perspectives -1
1:00:16 f-divergences
1:01:24 Why the f-divergence?
1:04:03 Statistical Machine Learning Perspective
1:04:48 Margin-Based Loss Functions
1:05:40 Estimation Based on a Convex Surrogate Loss
1:07:02 Some Theory for Surrogate Loss Functions
1:08:38 Outline -2
1:09:39 Setup -1
1:11:28 Profiling
1:13:44 Some Examples
1:14:19 Link -1
1:15:00 Conjugate Duality
1:15:49 Link -2
1:17:12 The Easy Direction
1:18:20 The Direction Has a Constructive Consequence -1
1:19:07 Example - Hellinger distance
1:20:45 Example - Variational distance
1:21:24 Example - Kullback-Leibler divergence
1:21:29 Bayes Consistency -1
_____________________
Source: videolectures.net/mlss09uk_jor...
This guy is so much better than LeBron why is there even a debate
At basketball, maybe. But I believe LeBron is still a better statistician.
Thanks, I'm sure quite a few people from Lex's podcast will be wanting to check this out :)
+1 from Lex's podcast!
I swing both ways. I've found it works.
Lex sent me here. Thanks for posting! The main Google hit for this video directs to a website so old that it expects the user to be on Windows 95. Video refuses to play on my machine.
You are very welcome! I figured many others would benefit from it being on youtube.
I was hoping to find a few "I'm just confused" but nah😅 everyone is solid in their outlook😂
Slides: mlg.eng.cam.ac.uk/mlss09/mlss_slides/Jordan_1.pdf
where are the slides??
I don’t understand the dilemma. Bayes theorem is a theorem that is trivially proved. There’s absolutely no question as to whether it’s correct or not.
Is the question about the accuracy of the chosen prior? If you have enough samples the prior doesn’t matter because you’ll converge on the truth either way. If you don’t have enough data for a frequentist convergence then isn’t the prior the best you can do?
Thank you for posting your Q. I'd like to answer your Q. Before that, i need to clarify, "What is the dilemma ?"
Who here in 2025
Currently April 2024. Captivating talk. Very real world.