IS = { zkontrolovano 14 Feb 2005 },
  UPDATE  = { 2005-01-07 },
  author =   {Hlav{\'a}{\v c}, V{\'a}clav and K{\'a}lal, Zden{\v e}k},
  title =   {Learning Finite Automaton from Noisy Observations -- A
                  Simple Instance of a Bidirectional Signal-to-symbol
                  Interface (The initial experiment)},
  institution =   {Center for Machine Perception, K13133 FEE Czech Technical
  address =   {Prague, Czech Republic},
  year =   {2004},
  month =   {December},
  type =   {Research Report},
  number =   {{CTU--CMP--2004--13}},
  issn =   {1213-2365},
  pages =   {33},
  figures =   {15},
  authorship =   {50-50},
  psurl =   { [HlavacKalalTR2004-13.pdf]},
  project =   {GACR 102/03/0440, CONEX GZ 45.535, IST-004176},
  annote =   {This report investigates the way how to learn the finite
                  automaton model of the activity observed in real world. The
                  related theory is reviewed, solution proposed and
                  experiments conducted. Learning finite automaton is similar
                  to learning a discrete Hidden Markov Model (HMM) using a
                  variant of EM algorithm. We used J. Dupa{\v c}'s discrete HMM
                  Toolbox in Matlab. The experimental part of this work deals
                  with learning HMM from a synthetic training set generated
                  from a known model. This approach provides us with ground
  keywords =   {Learning finite automaton, Hidden Markov Model, Machine