Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 8
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Discriminative Bernoulli HMMs for isolated handwritten word recognition

  • Giménez Pastor, Adrián
  • Andrés Ferrer, Jesús
  • Juan, Alfons
[EN] Bernoulli HMMs (BHMMs) have been successfully applied to handwritten text recognition (HTR) tasks such as continuous and isolated handwritten words. BHMMs belong to the generative model family and, hence, are usually trained by (joint) maximum likelihood estimation (MLE) by means of the Baum-Welch algorithm. Despite the good properties of the MLE criterion, there are better training criteria such as maximum mutual information (MM!). The MMI is the most widespread criterion to train discriminative models such as log-linear (or maximum entropy) models. Inspired by a BHMM classifier, in this work, a log-linear HMM (LLHMM) for binary data is proposed. The proposed model is proved to be equivalent to the BHMM classifier, and, in this way, a discriminative training framework for BHMM classifiers is defined. The behavior of the proposed discriminative training framework is deeply studied in a well known task of isolated word recognition, the RIMES database. (C) 2013 Elsevier B.V. All rights reserved., Work supported by the EC (FEDER/FSE) and the Spanish MEC/MICINN under the MIPRCV ‘‘Consolider Ingenio 2010’’ program (CSD2007-00018), iTrans2 (TIN2009-14511) and MITTRAL (TIN2009-14633-C03-01) projects. Also supported by the IST Programme of the European Community, under the PASCAL2 Network of Excellence, IST-2007-216886, and by the Spanish MITyC under the erudito.com (TSI-020110-2009-439).
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Combining Multiple Hypothesis Testing with Machine Learning Increases the Statistical Power of Genome-wide Association Studies

  • Mieth, Bettina
  • Rodríguez-Pérez, Juan Antonio
  • Morcillo-Suárez, Carlos
  • Farré, Xavier
  • Navarro, Arcadi
  • Müller, Klaus-Robert
Mieth, Bettina et al., The standard approach to the analysis of genome-wide association studies (GWAS) is based on testing each position in the genome individually for statistical significance of its association with the phenotype under investigation. To improve the analysis of GWAS, we propose a combination of machine learning and statistical testing that takes correlation structures within the set of SNPs under investigation in a mathematically well-controlled manner into account. The novel two-step algorithm, COMBI, first trains a support vector machine to determine a subset of candidate SNPs and then performs hypothesis tests for these SNPs together with an adequate threshold correction. Applying COMBI to data from a WTCCC study (2007) and measuring performance as replication by independent GWAS published within the 2008–2015 period, we show that our method outperforms ordinary raw p-value thresholding as well as other state-of-the-art methods. COMBI presents higher power and precision than the examined alternatives while yielding fewer false (i.e. non-replicated) and more true (i.e. replicated) discoveries when its results are validated on later GWAS studies. More than 80% of the discoveries made by COMBI upon WTCCC data have been validated by independent studies. Implementations of the COMBI method are available as a part of the GWASpi toolbox 2.0., EF acknowledges support from the advanced ERC grant (ERC-2011-AdG 295642-FEP) on the Foundation of Economic Preferences. MK, BM, and KRM were supported by the German National Science Foundation (DFG) under the grants MU 987/6-1 and RA 1894/1-1. TD and DS were supported by the German National Science Foundation (DFG) under the grants DI 1723/3-1 und SCHU 2828/2-1. GB and TS acknowledge support of the German National Science Foundation (DFG) under the research group grant FOR 1735. MK, DT, KRM, and GB acknowledge financial support by the FP7-ICT Programme of the European Community, under the PASCAL2 Network of Excellence. MK acknowledges a postdoctoral fellowship by the German Research Foundation (DFG), award KL 2698/2-1, and from the Federal Ministry of Science and Education (BMBF) awards 031L0023A and 031B0187B. AN acknowledges support from the Spanish Multiple Sclerosis Network (REEM), of the Instituto de Salud Carlos III (RD12/0032/0011), the Spanish National Institute for Bioinformatics (PT13/0001/0026) the Spanish Government Grant BFU2012-38236 and from FEDER. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 634143 (MedBioinformatics). MK and KRM were financially supported by the Ministry of Education, Science, and Technology, through the National Research Foundation of Korea under Grant R31-10008 (MK, KRM) and BK21 (KRM)., Peer reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Learning probabilistic automata : a study in state distinguishability

  • Balle Pigem, Borja de
  • Castro Rabal, Jorge|||0000-0002-1390-1313
  • Gavaldà Mestre, Ricard|||0000-0003-4736-7179
Known algorithms for learning PDFA can only be shown to run in time polynomial in the so-called distinguishability μ of the target machine, besides the number of states and the usual accuracy and confidence parameters. We show that the dependence on μ is necessary in the worst case for every algorithm whose structure resembles existing ones. As a technical tool, a new variant of Statistical Queries termed View the MathML source-queries is defined. We show how to simulate View the MathML source-queries using classical Statistical Queries and show that known PAC algorithms for learning PDFA are in fact statistical query algorithms. Our results include a lower bound: every algorithm to learn PDFA with queries using a reasonable tolerance must make Ω(1/μ1−c) queries for every c>0. Finally, an adaptive algorithm that PAC-learns w.r.t. another measure of complexity is described. This yields better efficiency in many cases, while retaining the same inevitable worst-case behavior. Our algorithm requires fewer input parameters than previously existing ones, and has a better sample bound., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Mining frequent closed rooted trees

  • Balcázar Navarro, José Luis|||0000-0003-4248-4528
  • Bifet Figuerol, Albert Carles
  • Lozano Boixadors, Antoni|||0000-0002-3633-063X
Many knowledge representation mechanisms are based on tree-like structures, thus symbolizing the fact that certain pieces of information are related in one sense or another. There exists a well-studied process of closure-based data mining in the itemset framework: we consider the extension of this process into trees. We focus mostly on the case where labels on the nodes are nonexistent or unreliable, and discuss algorithms for closurebased mining that only rely on the root of the tree and the link structure. We provide a notion of intersection that leads to a deeper understanding of the notion of support-based closure, in terms of an actual closure operator. We describe combinatorial characterizations and some properties of ordered trees, discuss their applicability to unordered trees, and rely on them to design efficient algorithms for mining frequent closed subtrees both in the ordered and the unordered settings. Empirical validations and comparisons with alternative algorithms are provided.
UPCommons. Portal del coneixement obert de la UPC
Publicación Artículo científico (JournalArticle)

Zipf's law for word frequencies: Word forms versus lemmas in long texts

  • Corral, Alvaro
  • Boleda Torrent, Gemma
  • Ferrer Cancho, Ramon|||0000-0002-7820-923X
Zipf's law is a fundamental paradigm in the statistics of written and spoken natural language as well as in other communication systems. We raise the question of the elementary units for which Zipf's law should hold in the most natural way, studying its validity for plain word forms and for the corresponding lemma forms. We analyze several long literary texts comprising four languages, with different levels of morphological complexity. In all cases Zipf's law is fulfilled, in the sense that a power-law distribution of word or lemma frequencies is valid for several orders of magnitude. We investigate the extent to which the word-lemma transformation preserves two parameters of Zipf's law: the exponent and the low-frequency cut-off. We are not able to demonstrate a strict invariance of the tail, as for a few texts both exponents deviate significantly, but we conclude that the exponents are very similar, despite the remarkavble transformation that going from words to lemmas represents, considerably affecting all ranges of frequencies. In contrast, the low-frequency cut-offs are less stable, tending to increase substantially after the transformation., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Learning probabilistic automata : a study in state distinguishability

  • Balle Pigem, Borja de
  • Castro Rabal, Jorge
  • Gavaldà Mestre, Ricard
Known algorithms for learning PDFA can only be shown to run in time polynomial in the so-called distinguishability μ of the target machine, besides the number of states and the usual accuracy and confidence parameters. We show that the dependence on μ is necessary in the worst case for every algorithm whose structure resembles existing ones. As a technical tool, a new variant of Statistical Queries termed View the MathML source-queries is defined. We show how to simulate View the MathML source-queries using classical Statistical Queries and show that known PAC algorithms for learning PDFA are in fact statistical query algorithms. Our results include a lower bound: every algorithm to learn PDFA with queries using a reasonable tolerance must make Ω(1/μ1−c) queries for every c>0. Finally, an adaptive algorithm that PAC-learns w.r.t. another measure of complexity is described. This yields better efficiency in many cases, while retaining the same inevitable worst-case behavior. Our algorithm requires fewer input parameters than previously existing ones, and has a better sample bound., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Mining frequent closed rooted trees

  • Balcázar Navarro, José Luis
  • Bifet Figuerol, Albert Carles
  • Lozano Boixadors, Antoni
Many knowledge representation mechanisms are based on tree-like structures, thus symbolizing the fact that certain pieces of information are related in one sense or another. There exists a well-studied process of closure-based data mining in the itemset framework: we consider the extension of this process into trees. We focus mostly on the case where labels on the nodes are nonexistent or unreliable, and discuss algorithms for closurebased mining that only rely on the root of the tree and the link structure. We provide a notion of intersection that leads to a deeper understanding of the notion of support-based closure, in terms of an actual closure operator. We describe combinatorial characterizations and some properties of ordered trees, discuss their applicability to unordered trees, and rely on them to design efficient algorithms for mining frequent closed subtrees both in the ordered and the unordered settings. Empirical validations and comparisons with alternative algorithms are provided., Postprint (author’s final draft)
Recercat. Dipósit de la Recerca de Catalunya
Publicación Artículo científico (JournalArticle)

Zipf's law for word frequencies: Word forms versus lemmas in long texts

  • Corral, Alvaro
  • Boleda Torrent, Gemma
  • Ferrer Cancho, Ramon
Zipf's law is a fundamental paradigm in the statistics of written and spoken natural language as well as in other communication systems. We raise the question of the elementary units for which Zipf's law should hold in the most natural way, studying its validity for plain word forms and for the corresponding lemma forms. We analyze several long literary texts comprising four languages, with different levels of morphological complexity. In all cases Zipf's law is fulfilled, in the sense that a power-law distribution of word or lemma frequencies is valid for several orders of magnitude. We investigate the extent to which the word-lemma transformation preserves two parameters of Zipf's law: the exponent and the low-frequency cut-off. We are not able to demonstrate a strict invariance of the tail, as for a few texts both exponents deviate significantly, but we conclude that the exponents are very similar, despite the remarkavble transformation that going from words to lemmas represents, considerably affecting all ranges of frequencies. In contrast, the low-frequency cut-offs are less stable, tending to increase substantially after the transformation., Peer Reviewed, Postprint (published version)