Loading...
Loading...

Go to the content (press return)

Spectral learning of weighted automata: a forward-backward perspective

Author
B. Balle; Carreras, X.; Luque, F. M.; Quattoni, A.J.
Type of activity
Journal article
Journal
Machine learning
Date of publication
2013-10-07
Number
October
First page
1
Last page
31
DOI
https://doi.org/10.1007/s10994-013-5416-x Open in new window
Project funding
Biological and Social Data Mining: Algorithms, Theory, and Implementations (TIN2011-27479-C04-03)
Cross-lingual Knowledge Extraction
Pattern Analysis, Statistical Modelling, and Computational Learning 2 (PASCAL2)
Repository
http://hdl.handle.net/2117/21075 Open in new window
URL
http://link.springer.com/article/10.1007%2Fs10994-013-5416-x Open in new window
Abstract
In recent years we have seen the development of efficient provably correct algorithms for learning Weighted Finite Automata (WFA). Most of these algorithms avoid the known hardness results by defining parameters beyond the number of states that can be used to quantify the complexity of learning automata under a particular distribution. One such class of methods are the so-called spectral algorithms that measure learning complexity in terms of the smallest singular value of some Hankel matrix. Ho...
Citation
Balle, B. [et al.]. Spectral learning of weighted automata: a forward-backward perspective. "Machine learning", 07 Octubre 2013, núm. October, p. 1-31.
Keywords
Spectral learning Weighted finite automata Dependency parsing
Group of research
GPLN - Natural Language Processing Group
LARCA - Laboratory of Relational Algorithmics, Complexity and Learnability

Participants

  • De Balle Pigem, Borja  (author)
  • Carreras Perez, Xavier  (author)
  • Luque, Franco M.  (author)
  • Quattoni, Ariadna Julieta  (author)