On the computational complexity of approximating distributions by probabilistic automata

@article{Abe1990OnTC,
  title={On the computational complexity of approximating distributions by probabilistic automata},
  author={Naoki Abe and Manfred K. Warmuth},
  journal={Machine Learning},
  year={1990},
  volume={9},
  pages={205-260}
}
We introduce a rigorous performance criterion for training algorithms for probabilistic automata (PAs) and hidden Markov models (HMMs), used extensively for speech recognition, and analyze the complexity of the training problem as a computational problem. The PA training problem is the problem of approximating an arbitrary, unknown source distribution by distributions generated by a PA. We investigate the following question about this important, well-studied problem: Does there exist… CONTINUE READING

Citations

Publications citing this paper.
SHOWING 1-10 OF 124 CITATIONS

On learning finite-state quantum sources

  • Quantum Information & Computation
  • 2009
VIEW 7 EXCERPTS
CITES METHODS
HIGHLY INFLUENCED

Pattern classification via unsupervised learners

VIEW 4 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

PAC-learnability of Probabilistic Deterministic Finite State Automata

  • J. Mach. Learn. Res.
  • 2004
VIEW 4 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Learning Geometrically-Constrained Hidden Markov Models for Robot Navigation: Bridging the Topological-Geometrical Gap

H. Shatkay, L. P. Kaelbling
  • J. Artif. Intell. Res.
  • 2002
VIEW 5 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

Designing Statistical Language Learners: Experiments on Noun Compounds

  • ArXiv
  • 1996
VIEW 4 EXCERPTS
CITES BACKGROUND
HIGHLY INFLUENCED

On the Learnability and Usage of Acyclic Probabilistic Finite Automata

VIEW 4 EXCERPTS
CITES BACKGROUND & METHODS
HIGHLY INFLUENCED

FILTER CITATIONS BY YEAR

1989
2018

CITATION STATISTICS

  • 16 Highly Influenced Citations

References

Publications referenced by this paper.
SHOWING 1-10 OF 12 REFERENCES

Minimum consistent 2-state DFA problem is NP-complete

D. Angluin
  • 1989
VIEW 7 EXCERPTS
HIGHLY INFLUENTIAL

Complexity of Automaton Identification from Given Data

  • Information and Control
  • 1978
VIEW 4 EXCERPTS
HIGHLY INFLUENTIAL

On the Complexity of Minimum Inference of Regular Sets

  • Information and Control
  • 1978
VIEW 5 EXCERPTS
HIGHLY INFLUENTIAL

An introduction to the application of the theory of the probabilistic functions of a Markov process to automatic speech recognition

San Mateo, L. R. Rabiner, M. M. Sondhi
  • 1988