[mlpack-svn] [MLPACK] #315: set initial state probabilities for hmm
MLPACK Trac
trac at coffeetalk-1.cc.gatech.edu
Sun Jan 12 13:12:50 EST 2014
#315: set initial state probabilities for hmm
--------------------------+-------------------------------------------------
Reporter: fleischhauf | Owner:
Type: enhancement | Status: reopened
Priority: minor | Milestone:
Component: mlpack | Resolution:
Keywords: | Blocking:
Blocked By: |
--------------------------+-------------------------------------------------
Changes (by fleischhauf):
* status: closed => reopened
* resolution: fixed =>
Comment:
Hey Ryan,
thats not what I meant,
usually an hidden markov model is parameterized by:
Transition matrix A,
Observation probabilities B,
initial state probabilities \pi.
like described in
here:[http://courses.media.mit.edu/2010fall/mas622j/ProblemSets/ps4/tutorial.pdf]
or here: [http://people.sabanciuniv.edu/~berrin/cs512/reading/rabiner-
tutorial-on-hmm.pdf]
or are the initial state probabilities somehow part of the transition
matrix in the constructor ?
--
Ticket URL: <http://trac.research.cc.gatech.edu/fastlab/ticket/315#comment:2>
MLPACK <www.fast-lab.org>
MLPACK is an intuitive, fast, and scalable C++ machine learning library developed at Georgia Tech.
More information about the mlpack-svn
mailing list