Search found 22 matches

by snowave
Fri Dec 11, 2020 3:37 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

Re: How log likelihood value is calculated in EM?

Hi Experts, I use Genie to do EM learning and switch between JT inference and EPIS sampling algorithms. Why the parameter learned are exactly the same (even for the decimals)? Though the marginal results are different. It seems like the parameter learning is tied to an exact method. Only at the marg...
by snowave
Wed Nov 04, 2020 9:44 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

Re: How log likelihood value is calculated in EM?

Thanks a lot.
by snowave
Wed Nov 04, 2020 8:31 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

Re: How log likelihood value is calculated in EM?

Hi experts, if the model has high tree-width, i.e., the inference querying posterior joint distributions can't be performed by exact methods like JT due to memory constraint, what approximate algorithm is used in EM learning? Are they relevance based decomposition and/or sampling methods? Many Thanks
by snowave
Tue Jul 28, 2020 9:38 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

Re: How log likelihood value is calculated in EM?

cool, very much appreciated.
by snowave
Tue Jul 28, 2020 8:52 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

Re: How log likelihood value is calculated in EM?

I think I understand the chain rule algorithm.

But I can't understand the "Relevant decomposition of JT". Is there a link or more detailed explanation for that? appreciate a lot.
by snowave
Tue Jul 28, 2020 8:26 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

Re: How log likelihood value is calculated in EM?

many thanks, so the log p(x)=log (product of cliques/product of sepsets)?
Could you please explain a bit more on chain rule calculation of the log p(x)?

best
by snowave
Tue Jul 28, 2020 6:37 am
Forum: SMILE
Topic: How log likelihood value is calculated in EM?
Replies: 11
Views: 8673

How log likelihood value is calculated in EM?

Hi,

Can anyone explain how the log likelihood value in EM learning is calculated in Genie? Suppose the underline inference algorithm is Junction Tree.
In my understanding the log P(x) is calculated by multiplying and dividing cliques, is that correct?

many thanks
by snowave
Mon Sep 03, 2018 12:31 pm
Forum: SMILE
Topic: EM algorithm for discrete model in GeNIe
Replies: 2
Views: 4547

Re: EM algorithm for discrete model in GeNIe

Thanks, so for the discrete EM we are actually filling in the missing data with expected values and then use the complete data to learn parameters.

Peng
by snowave
Sat Sep 01, 2018 9:14 am
Forum: SMILE
Topic: EM algorithm for discrete model in GeNIe
Replies: 2
Views: 4547

EM algorithm for discrete model in GeNIe

Hi, it seems most materials available online are explaining the EM for GMM models, i.e. generate expected sufficient statistics for Gaussians.
But how EM works on discrete models? Especially the algorithm used in GeNIe? Can I ask for a reference or simple explanations? Thanks a lot.

Peng
by snowave
Thu Mar 16, 2017 9:33 am
Forum: SMILE
Topic: SMILE and GENIE structure learning
Replies: 3
Views: 4726

Re: SMILE and GENIE structure learning

It just runs for a long time. Thanks. I changed from debug mode to release mode.
by snowave
Thu Mar 16, 2017 4:13 am
Forum: SMILE
Topic: Error message when update net beliefs
Replies: 2
Views: 3900

Re: Error message when update net beliefs

many thanks it works.
by snowave
Thu Mar 16, 2017 2:33 am
Forum: SMILE
Topic: SMILE and GENIE structure learning
Replies: 3
Views: 4726

Re: SMILE and GENIE structure learning

I'm using VS2015
by snowave
Wed Mar 15, 2017 8:26 am
Forum: SMILE
Topic: Error message when update net beliefs
Replies: 2
Views: 3900

Error message when update net beliefs

I've leaned network structure with 2100 variables using gtt. When I update the network there is some errors:
Error (-42): UpdateBeliefs failed - network structure is too complex.

Could you please help sovle this?

many thanks
snowave
by snowave
Wed Mar 15, 2017 4:34 am
Forum: SMILE
Topic: SMILE and GENIE structure learning
Replies: 3
Views: 4726

SMILE and GENIE structure learning

Hi, I have a relatively large dataset with 2500 features. I use SMILE and GENIE to perform greedy thick thinning structure learning. SMILE (in C++) seems going to run infinitely long without a response. GENIE, when using max num of parents=3, runs for 7 minutes and pop an error. If instead, of using...
by snowave
Tue Jan 17, 2017 11:49 am
Forum: SMILE
Topic: jsmile EM
Replies: 1
Views: 3150

jsmile EM

Hi, I'm trying to use jsmile EM functionality. Is there an example/ tutorial for that?
I have problems to use the function learn(), which has a parameter DataMatch[] matching to pass in. How to set DataMatch [] matching?

Thanks

snowave