Search found 19 matches

by musicpxg
Tue Oct 07, 2014 6:06 pm
Forum: SMILE
Topic: In EM learning, how to keep the right relationships
Replies: 2
Views: 2811

Re: In EM learning, how to keep the right relationships

mark wrote:After running EM you can swap the b0 and b1 state names.
It's a good idea.

I found there is a function network.setOutcomeId(), I think it can be a solution.

Thank you!
by musicpxg
Tue Oct 07, 2014 1:52 pm
Forum: SMILE
Topic: In EM learning, how to keep the right relationships
Replies: 2
Views: 2811

In EM learning, how to keep the right relationships

Hi, I will show my question by an example: For instance, I have three nodes, A, B, C C is child of B, B is child of A. A has two states, a0,a1 B has two states, b0,b1 C has two states, c0,c1 B is an invisible node in the training data set. Only data for A and C are available. a0 has strong relations...
by musicpxg
Mon Oct 06, 2014 9:51 pm
Forum: SMILE
Topic: Consumed problem about EM learning
Replies: 4
Views: 6164

Re: Consumed problem about EM learning

Thank you! But what is EM.setRelevance(true) means?

Best wishes,
Yang
by musicpxg
Mon Oct 06, 2014 9:41 pm
Forum: SMILE
Topic: Consumed problem about EM learning
Replies: 4
Views: 6164

Re: Consumed problem about EM learning

I tried to run the code 5 times, 1st, 2nd, 4 and 5 works. The 3rd one reported such an error.

I tried again, the error occurred randomly.
by musicpxg
Mon Oct 06, 2014 9:38 pm
Forum: SMILE
Topic: Consumed problem about EM learning
Replies: 4
Views: 6164

Consumed problem about EM learning

Hi, I found that if I repeated to run my code, an error occurred randomly as follow: smile.SMILEException: SMILE error -1 in function Learn. Logged information: em: please run with relevance enabled My code is based on java, and it is a smile main(), each time, the programe automatically close after...
by musicpxg
Fri Oct 03, 2014 9:48 am
Forum: SMILE
Topic: How to get EM log-likelihood
Replies: 2
Views: 2679

Re: How to get EM log-likelihood

Can anybody help me?
by musicpxg
Thu Oct 02, 2014 6:44 am
Forum: SMILE
Topic: How to get EM log-likelihood
Replies: 2
Views: 2679

How to get EM log-likelihood

Hi, I am using the EM algorithm to lean BBN. Could anybody let me know how to get the EM log-likelihood? I tried EM em = new EM(); ...... em.learn(data, network, data.matchNetwork(network)); double likehood = em.getLastScore(); However, the result is always 0.0 What's wrong? I am using java. Thank y...
by musicpxg
Fri Sep 26, 2014 6:24 am
Forum: SMILE
Topic: Can set value to Dataset from array, list, etc?
Replies: 4
Views: 3120

Re: Can set value to Dataset from array, list, etc?

I find there is a wrapper for Java to generate .jar. Can I add a new function in the wrapper and then regenerate the smile.jar? You can do that, but you'd loose your changes when new version of jSMILE is posted. Why don't you create a class derived from DataSet or an utility class or set of classes...
by musicpxg
Thu Sep 25, 2014 1:19 pm
Forum: SMILE
Topic: Can set value to Dataset from array, list, etc?
Replies: 4
Views: 3120

Re: Can set value to Dataset from array, list, etc?

There's no direct support for Java collections, but you can build your dataset with the following methods defined in the DataSet class: addIntVariable, addFloatVariable, addEmptyRecord, setInt, setFloat, setMissing. Define the structure of your dataset with addXXXVariable calls. To add rows, call a...
by musicpxg
Thu Sep 25, 2014 9:03 am
Forum: SMILE
Topic: Can set value to Dataset from array, list, etc?
Replies: 4
Views: 3120

Can set value to Dataset from array, list, etc?

Hi,

I find there seems only a way to read data to Dataset, i.e. readFile(). Is it possible to set values to Dataset from Arraylist, etc? I do not like to save the arraylist to hard disk as .text and then readed by Dataset.readFile();

Thank you!
Best wishes,
Yang
by musicpxg
Fri May 23, 2014 1:41 pm
Forum: SMILE
Topic: The differences of the learning results
Replies: 5
Views: 3748

Re: The differences of the learning results

Hi shooltz,

What the meaning of "setSeed"? Can I just discard the lines, as follow

SEED_VALUE = 0;
em.setSeed(SEED_VALUE);

Thank you!
Have a nice Weekend!

Best wishes,
Yang
by musicpxg
Tue Apr 29, 2014 11:19 am
Forum: SMILE
Topic: The differences of the learning results
Replies: 5
Views: 3748

Re: The differences of the learning results

Hi,

The SEED_VALUE is 0.

Thanks!

Yang
by musicpxg
Fri Apr 25, 2014 4:07 pm
Forum: SMILE
Topic: The differences of the learning results
Replies: 5
Views: 3748

The differences of the learning results

Hi, I found the learning results from the Smile were different from that using Genie. There are only three nodes, A, B, H. An arc from H to A and B respectively. The CPT from H to A is H h0 h1 a0 0.9215585808924782 0.1103564750941653 a1 0.0784414191075218 0.8896435249058347 It said: EM log-likelihoo...
by musicpxg
Wed Apr 23, 2014 7:45 pm
Forum: SMILE
Topic: How to combine the prior knowledge during learning
Replies: 1
Views: 2656

How to combine the prior knowledge during learning

Hi, I already designed the structure of a network. I want to learn the parameters. There are questions as follows: There are no training data for a root node in my network. I have prior experience to assign its value. So, how can I hand this problem when I use the EM algorithm? Is it a good idea to ...
by musicpxg
Wed Apr 16, 2014 6:59 am
Forum: SMILE
Topic: How to combine two BBNs
Replies: 2
Views: 3026

Re: How to combine two BBNs

Alternatively, you can use some XML API to join two .xdsl files into one temporary file (assuming no node id conflicts in net_a and net_b). After loading the join file you would add the arc. Yes, this is a very smart idea! Thanks a lot to shooltz! You always response fast and professionally. Yang