- 115 Views
- Uploaded on

Download Presentation
## PowerPoint Slideshow about 'Using MCMC' - edison

**An Image/Link below is provided (as is) to download presentation**

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -

Presentation Transcript

Using MCMC

- Separating MCMC from Bayesian Inference?
- Line fitting revisited
- A toy equaliser problem
- Some lessons
- A problem in film restoration/retouching

[Bayesian Inference]

Try to see if you can

integrate out nuisances

Derive the Posterior

MCMC is (just) a toolChoose a Model

Identify Parameters

Need better model

If solution not ok

Use MCMC

Solve Deterministically

Direct, CG, Steepest Descent etc

Manipulate Random Samples

To get one answer

(if you want)

Gives you one answer

Can always design single parameter-at-a-time schemes. So iterations can be very low complexity

Simple iterations = long convergence

Gives you a picture of alternate answers

Do you really need alternate answers?

Will always allow you to get to “best” solution

Iterations can be high complexity?

Convergence can be rapid (e.g. CG) for well defined problems

Gives you just one answer

Can give local minimum for non-linear problems

Good, BadOthers

MCMC

Ugly iterations can be very low complexity

- To solve your problem you need a good model
- MCMC is not really going to help you if you have the wrong model
- MCMC suited to BIG problems: but what is BIG really?
- E.g. Exhaustive search for motion estimation is possible in real time (TV rates) in hardware: why bother with other things? (an exaggeration … but interesting nevertheless)

Line Fitting (again) iterations can be very low complexity

Needs Latex

Observed Data

Actual Line

Initial Guess

Typical Results iterations can be very low complexity

See Matlab demo

Nice Convergence

because we can draw samples directly

Watch out iterations can be very low complexity

- All random number generators are not created equal
- (See NR)
- Harder problems require longer runs (of course)
- Sometimes hard to get all bugs out because its all a random search anyway

Blind (?) Equalisation iterations can be very low complexity

Noise

Signal

2nd Order

All pole System

Rec’d Signal

Identify the system coefficients

AND recover the original signal

Comms, Deblurring, Overshoot Cancellation

Equaliser Problem iterations can be very low complexity

Now more latex

Direct numerical sampling iterations can be very low complexity

P(1) = 0.3, p(2) = 0.25,

P(3) = 0.2, p(4) = 0.25

0

1

0.3

2

0.25

3

0.2

0.75

4

0.25

1

Number line

interpretation

71 points evaluated

Gibbs sampler 1 iterations can be very low complexity(equaliser1.m)

Back to Latex

Typical

Estimated System

Actual System

X 20 !

300 iterations

Gibbs Sampler II iterations can be very low complexity(equaliser2.m)Using Filter Bank (system choices)

30 filters

Samples from filter bank iterations can be very low complexity

Samples of signal parameter iterations can be very low complexity

System Estimate iterations can be very low complexity

Equalised signal iterations can be very low complexity

Lessons iterations can be very low complexity

- Gibbs sampler takes big problems and breaks them into lots of small ones
- Spotting the functional form of a known p.d.f. is a useful skill. Books help.
- If all else fails, can always sample directly
- MCMC does not necessarily solve your problem. Good priors, better models are still important
- Deterministic/Stochastic Hybrid mix is v. useful

Download Presentation

Connecting to Server..