- Research Grants
|Support type:||Scholarships in Brazil - Doctorate|
|Effective date (Start):||April 01, 2009|
|Effective date (End):||February 29, 2012|
|Field of knowledge:||Physical Sciences and Mathematics - Probability and Statistics|
|Principal Investigator:||Nancy Lopes Garcia|
|Home Institution:||Instituto de Matemática, Estatística e Computação Científica (IMECC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil|
The goal of this project is to propose inferencial methods, specially hypothesis testing, for variable length Markov chains towards application to linguistic data. This work is closely related to a big research project that has been developing along the last years. One of the main projects related to this subject is the Projeto Temático "Comportamento estocástico, fenômenos críticos e identificação de padrões rítmicos nas línguas naturais", Processo FAPESP 03/09930-9.Several estimation methods are described in the literature. However, none of these articles propose hypothesis testing methods. This is due to the fact that the estimators are stochastic context trees and there is no obvious concept for variablity of these tress. We will treat this subject using resampling methods, specially one method based on the bootstrap of independent blocks.