cameron.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
decltype_me has joined #mlpack
billLiu has joined #mlpack
< billLiu>
Hey, guys. When I use the GMM module, I found some problem. The probability calculated is not log scaled which make it very possible to overflow when the dataset or the number of mixture is very large. Because the probability that assign to one component tends to zero or one. Is it a bug?
< billLiu>
The numeric problem is serious in my situation. I think we should calculate the loglikelihood instead. Is that right
decltype_me has quit [Quit: Konversation terminated!]