Learning nonsingular phylogenies and hidden Markov models
Elchanan Mossel and Sébastien Roch
Source: Ann. Appl. Probab. Volume 16, Number 2 (2006), 583-614.
Abstract
In this paper we study the problem of learning phylogenies and hidden Markov models. We call a Markov model nonsingular if all transition matrices have determinants bounded away from 0 (and 1). We highlight the role of the nonsingularity condition for the learning problem. Learning hidden Markov models without the nonsingularity condition is at least as hard as learning parity with noise, a well-known learning problem conjectured to be computationally hard. On the other hand, we give a polynomial-time algorithm for learning nonsingular phylogenies and hidden Markov models.
Primary Subjects: 60J10, 60J20, 68T05, 92B10
Keywords: Hidden Markov models; evolutionary trees; phylogenetic reconstruction; PAC learning
Full-text: Open access
Links and Identifiers
Permanent link to this document: http://projecteuclid.org/euclid.aoap/1151592244
Digital Object Identifier: doi:10.1214/105051606000000024
Zentralblatt MATH identifier:
1137.60034
References
Abe, N. and Warmuth, N. K. (1992). On the computational complexity of approximating probability distributions by probabilistic automata. Machine Learning 9 205--260.
Ambainis, A., Desper, R., Farach, M. and Kannan, S. (1997). Nearly tight bounds on the learnability of evolution. In Proceedings of the 38th Annual Symposium on Foundations of Computer Science 524--533. IEEE Computer Society, Washington, DC.
Blum, A., Furst, M., Jackson, J., Kearns, M., Mansour, Y. and Rudich, S. (1994). Weakly learning DNF and characterizing statistical query learning using fourier analysis. In Proceedings of the Twenty-Sixth Annual ACM Symposium on Theory of Computing 253--262. ACM Press, New York.
Blum, A., Kalai, A. and Wasserman, H. (2003). Noise-tolerant learning, the parity problem, and the statistical query model. J. ACM 50 506--519.
Mathematical Reviews (MathSciNet):
MR2146884
Digital Object Identifier: doi:10.1145/792538.792543
Chang, J. T. (1996). Full reconstruction of Markov models on evolutionary trees: Identifiability and consistency. Math. Biosci. 137 51--73.
Mathematical Reviews (MathSciNet):
MR1410044
Digital Object Identifier: doi:10.1016/S0025-5564(96)00075-2
Chor, B. and Tuller, T. (2005). Maximum likelihood of evolutionary trees is hard. Proceedings of Research in Computational Molecular Biology: 9th Annual International Conference. Lecture Notes in Comput. Sci. 3500 296--310. Springer, Berlin.
Mathematical Reviews (MathSciNet):
MR2304676
Cryan, M., Goldberg, L. A. and Goldberg, P. W. (2002). Evolutionary trees can be learned in polynomial time in the two-state general Markov model. SIAM J. Comput. 31 375--397.
Mathematical Reviews (MathSciNet):
MR1861281
Digital Object Identifier: doi:10.1137/S0097539798342496
Durbin, R., Eddy, S., Krogh, A. and Mitchison, G. (1998). Biological Sequence Analysis: Probabilistic Models of Proteins and Nucleic Acids. Cambridge Univ. Press.
Durrett, R. (1996). Probability: Theory and Examples, 2nd ed. Duxbury, Belmont, CA.
Mathematical Reviews (MathSciNet):
MR1609153
Erdos, P. L., Steel, M., Szekely, L. and Warnow, T. (1997). A few logs suffice to build (almost) all trees. I. Random Structures Algorithms 14 153--184.
Mathematical Reviews (MathSciNet):
MR1667319
Erdos, P. L., Steel, M. A., Szekely, L. A. and Warnow, T. J. (1999). A few logs suffice to build (almost) all trees. II. Theoret. Comput. Sci. 221 77--118.
Mathematical Reviews (MathSciNet):
MR1700821
Digital Object Identifier: doi:10.1016/S0304-3975(99)00028-6
Felsenstein, J. (2004). Inferring Phylogenies. Sinauer, New York.
Farach, M. and Kannan, S. (1999). Efficient algorithms for inverting evolution. J. ACM 46 437--449.
Mathematical Reviews (MathSciNet):
MR1812126
Digital Object Identifier: doi:10.1145/320211.320212
Feldman, J., O'Donnell, R. and Servedio, R. (2005). Learning mixtures of product distributions over discrete domains. In Proceedings of 46th Symposium on Foundations of Computer Science 501--510. IEEE Computer Society.
Felsenstein, J. (1978). Cases in which parsimony or compatibility methods will be positively misleading. Systematic Zoology 27 410--410.
Garey, M. R. and Johnson, D. S. (1979). Computers and Intractability: A Guide to the Theory of NP-Completeness. Freeman, San Francisco, CA.
Mathematical Reviews (MathSciNet):
MR0519066
Zentralblatt MATH:
0411.68039
Golub, G. H. and Van Loan, C. H. (1996). Matrix Computations, 3rd ed. Johns Hopkins Univ. Press.
Mathematical Reviews (MathSciNet):
MR1417720
Zentralblatt MATH:
0865.65009
Graham, R. L. and Foulds, L. R. (1982). Unlikelihood that minimal phylogenies for a realistic biological study can be constructed in reasonable computational time. Math. Biosci. 60 133--142.
Mathematical Reviews (MathSciNet):
MR0678718
Digital Object Identifier: doi:10.1016/0025-5564(82)90125-0
Helmbold, D., Sloan, R. and Warmuth, M. (1992). Learning integer lattices. SIAM J. Comput. 21 240--266.
Mathematical Reviews (MathSciNet):
MR1154523
Digital Object Identifier: doi:10.1137/0221019
Horn, R. A. and Johnson, C. R. (1985). Matrix Analysis. Cambridge Univ. Press.
Mathematical Reviews (MathSciNet):
MR0832183
Zentralblatt MATH:
0576.15001
Kearns, M. J. (1998). Efficient noise-tolerant learning from statistical queries. J. ACM 45 983--1006.
Mathematical Reviews (MathSciNet):
MR1678849
Digital Object Identifier: doi:10.1145/293347.293351
Kearns, M. J., Mansour, Y., Ron, D., Rubinfeld, R., Schapire, R. E. and Sellie, L. (1994). On the learnability of discrete distributions. In Proceedings of the Twenty-Sixth Annual ACM Symposium on Theory of Computing 273--282. ACM Press, New York.
Kearns, M. J. and Vazirani, U. V. (1994). An Introduction to Computational Learning Theory. MIT Press.
Mathematical Reviews (MathSciNet):
MR1331838
Lyngs, R. B. and Pedersen, C. N. S. (2001). Complexity of comparing hidden Markov models. Proceedings of Algorithms and Computation, 12th International Symposium. Lecture Notes in Comput. Sci. 2223 416--428. Springer, Berlin.
Mathematical Reviews (MathSciNet):
MR1917761
Zentralblatt MATH:
1077.68654
Digital Object Identifier: doi:10.1007/3-540-45678-3_36
Mossel, E. (2003). On the impossibility of reconstructing ancestral data and phylogenies. J. Comput. Biol. 10 669--678.
Mossel, E. (2004). Distorted metrics on trees and phylogenetic forests. Preprint. Available at arxiv.org/math.CO/0403508.
Mathematical Reviews (MathSciNet):
MR2056226
Zentralblatt MATH:
1066.94006
Rabiner, L. R. (1989). A tutorial on hidden Markov models and selected applications in speech recognition. Proceedings of the IEEE 77 257--286.
Rice, K. and Warnow, T. (1997). Parsimony is hard to beat! Proceedings of the Third Annual International Conference on Computing and Combinatorics. Lecture Notes in Comput. Sci. 1276 124--133. Springer, Berlin.
Mathematical Reviews (MathSciNet):
MR1616310
Digital Object Identifier: doi:10.1007/BFb0045079
Roch, S. (2006). A short proof that phylogenetic tree reconstruction by maximum likelihood is hard. In IEEE/ACM Transactions on Computational Biology and Bioinformatics 3 92--94.
Semple, C. and Steel, M. (2003). Phylogenetics. Oxford Univ. Press.
Mathematical Reviews (MathSciNet):
MR2060009
Zentralblatt MATH:
1043.92026
Steel, M. (1994). Recovering a tree from the leaf colourations it generates under a Markov model. Appl. Math. Lett. 7 19--24.
Mathematical Reviews (MathSciNet):
MR1350139
Digital Object Identifier: doi:10.1016/0893-9659(94)90024-8
Valiant, L. G. (1984). A theory of the learnable. Comm. ACM 27 1134--1142.
The Annals of Applied Probability
