Search found 20 matches
- Wed Jan 04, 2017 5:30 pm
- Forum: SMILE
- Topic: Learned HMM parameters(by EM) different from BNT toolbox
- Replies: 3
- Views: 5386
Re: Learned HMM parameters(by EM) different from BNT toolbox
Dear Mark, Thank you so much for your quick reply! Following your advice, I did smaller steps to debug, and finally found out that I made an error: I shouldn't prepare all HMMs data in one single file, but should prepare one data file for one HMM instead. By doing in my previous way, different HMMs ...
- Tue Jan 03, 2017 2:43 pm
- Forum: SMILE
- Topic: Learned HMM parameters(by EM) different from BNT toolbox
- Replies: 3
- Views: 5386
Learned HMM parameters(by EM) different from BNT toolbox
Hi, dear SMILE creators, I tried to learn parameters of Hidden Markov Models using jsmile's EM class, and end up having quite different parameters from those from BNT(Bayes Net Toolbox for Matlab), given that I set the initial values of the parameters the same in both toolboxes. The possible reasons...
- Fri Oct 28, 2016 7:26 pm
- Forum: SMILE
- Topic: Create the DBN structure by JSMILE methods
- Replies: 2
- Views: 6032
Re: Create the DBN structure by JSMILE methods
Thank you so much!!! Now I can use JSMILE methods to create a network same as the the one from the tutorial :) In case it is useful to others, I am posting the code after fixing the problem: Network net = new Network(); String[] nodes = { "A", "B", "C" }; int nbSlices =...
- Wed Oct 19, 2016 2:25 am
- Forum: SMILE
- Topic: Create the DBN structure by JSMILE methods
- Replies: 2
- Views: 6032
Create the DBN structure by JSMILE methods
Hi, SMILE creators, Thank you so much for your time! I tried to create a DBN structure file using JSMILE's methods rather than "manual" coding (after successfully running em.learn of the DBN with the given structure file from the tutorial). I tried to create it in a "roll" way in...
- Tue Oct 18, 2016 7:32 pm
- Forum: SMILE
- Topic: Output posterior probabilities for latent variables (and missing values) after EM converges on training set
- Replies: 2
- Views: 5998
Output posterior probabilities for latent variables (and missing values) after EM converges on training set
Hi, SMILE creators, I am currently using JSMILE's em.learn to learn the parameters of a DBM with latent variables and missing values for observables. I want to check the posterior probabilities of the latent variables and observables' missing values for each sample (e.g., I have 20 samples correspon...