Bootstrap and model selection for stochastic chains with memory of variable length
Recurrent neural networks for classification of proteins into families and compari...
![]() | |
Author(s): |
David Henriques da Matta
Total Authors: 1
|
Document type: | Master's Dissertation |
Press: | Campinas, SP. |
Institution: | Universidade Estadual de Campinas (UNICAMP). Instituto de Matemática, Estatística e Computação Científica |
Defense date: | 2008-03-25 |
Examining board members: |
Nancy Lopes Garcia;
Jefferson Antonio Galves;
Jesus Enrique Garcia
|
Advisor: | Nancy Lopes Garcia |
Abstract | |
In this project, we focus our studies on the question of finding statistical evidences in detecting rhythm in written texts by presenting probabilistic tools that allow us to discriminate Brazilian and Portuguese texts. To achieve such goals, we some present theoretical and practical results in modeling, resampling and estimation of variable length Markov Chains. More over in the part, we propose a new method of resampling for data sets with a renewal point (AU) | |
FAPESP's process: | 05/03434-5 - Estimation of algorithms for variable length Markov chains |
Grantee: | David Henriques da Matta |
Support Opportunities: | Scholarships in Brazil - Master |