[mlpack] GMM result in NaN matrices

Jaelon joery_vn at hotmail.com
Wed Apr 10 06:06:12 EDT 2013


Hey,

I am currently working with GMM and it worked for two models with decent 
results up to now.
However when attempting to train for a third model I get mean, 
covariance and transmission matrices filled with NaN.
My input data has 15 dimensions consisting of values that range between 
0 and 4.
My guess is that excessive appearance of the value zero is causing the 
problem and thus I tried adding a very small value to every datapoint, 
but this did not solve the problem.

I am training 7 models with data like explained above. 2 of them always 
give me NaN matrices and take way longer to train so something is going 
wrong with the training.
Could you give me an possible reason for this problem and/or point me in 
the direction to solving this, I have been stuck several hours on this now).

PS. I updated from the repository about 6hours ago (version r14885)

With many thanks,
Joery


More information about the mlpack mailing list