An introduction to importance sampling - optimal importance distributions

  Рет қаралды 15,141

Ben Lambert

Ben Lambert

Күн бұрын

Пікірлер: 13
@Neural-k2z
@Neural-k2z 6 жыл бұрын
At 2:47, for the continuous case, isn't there an f(x) missing?
@SpartacanUsuals
@SpartacanUsuals 6 жыл бұрын
Yes, good spot. There should be a f(x) after the fraction. Sorry for any confusion this causes people! Best, Ben
@FalguniDasShuvo
@FalguniDasShuvo 4 жыл бұрын
@@SpartacanUsuals You could give it a quick fix by just popping up some text during the video saying there will be an additional f(x).
@marcoponts8942
@marcoponts8942 4 жыл бұрын
At 6:56, shouldn't there also be a multiplication with the function you divide by? Because the mean E_g[x] should only be sum over x*g(x), or should it not?
@quao8194
@quao8194 4 жыл бұрын
It's very helpful! Thank you very much for making this.
@karthikponnupillai8923
@karthikponnupillai8923 3 жыл бұрын
please guide me for this question: Suppose that you have run a Monte Carlo analysis (m samples) and wish to cut the standard deviation in half. How many samples do you need to run?
@stefanoskechagias1736
@stefanoskechagias1736 5 жыл бұрын
very nice! Maybe follow up with video on approximating likelihoods plus some resampling to avoid degeneracy.
@davidsewell4999
@davidsewell4999 6 жыл бұрын
Just curious if you use mathematica in your day to day workflow? The graphics / animations you have shown in mathematica seem like they could be really helpful and it requires a fair bit of work to do the same thing in something like python.
@SpartacanUsuals
@SpartacanUsuals 6 жыл бұрын
Hi, thanks for your comment. Yes, it is by far the most important tool in my day to day work (R is, perhaps, second). There's a bit of a learning curve when you start with Mathematica. But the ease with which you can create interactive graphics and animations really leaves Matlab in the dust, in my view. Best, Ben
@davidsewell4999
@davidsewell4999 6 жыл бұрын
Ben, Thanks for the response. Ok interesting well I may have to give it a go. I use python at work for doing different machine learning tasks but being able to create animations as you have done is very helpful for understanding different concepts. Anyways , great series and I am enjoying the book so far as well!
@costantinodibello8027
@costantinodibello8027 3 жыл бұрын
Very nice explanation! Can someone explain how to make that funny animation with Mathematica?
@omridrori3286
@omridrori3286 3 жыл бұрын
Hey can you give link for the animation so i can also play it myself?
@royapar7050
@royapar7050 Жыл бұрын
How we choose h(x)?
An introduction to importance sampling
14:19
Ben Lambert
Рет қаралды 59 М.
BAYGUYSTAN | 1 СЕРИЯ | bayGUYS
36:55
bayGUYS
Рет қаралды 1,9 МЛН
Quando eu quero Sushi (sem desperdiçar) 🍣
00:26
Los Wagners
Рет қаралды 15 МЛН
Rejection Sampling + R Demo
13:28
math et al
Рет қаралды 25 М.
Monte Carlo Simulation
10:06
MarbleScience
Рет қаралды 1,5 МЛН
1. Introduction to Statistics
1:18:03
MIT OpenCourseWare
Рет қаралды 2 МЛН
Importance Sampling
12:46
Mutual Information
Рет қаралды 67 М.
An introduction to the Random Walk Metropolis algorithm
11:28
Ben Lambert
Рет қаралды 62 М.
An introduction to Gibbs sampling
18:58
Ben Lambert
Рет қаралды 82 М.
Accept-Reject Sampling : Data Science Concepts
17:49
ritvikmath
Рет қаралды 72 М.
But what is the Central Limit Theorem?
31:15
3Blue1Brown
Рет қаралды 3,7 МЛН
The intuition behind the Hamiltonian Monte Carlo algorithm
32:09
Ben Lambert
Рет қаралды 60 М.