Optimal black-box sequential searching

=== This dissertation constructs optimal root-searching and aximumsearching algorithms in a statistical sense and compares the statistically optimal strategies to the already known mini-maximal strategies. In order to construct the so called statistical method, new results in the eld of probability...

Full description

Bibliographic Details
Main Author: Ivo Fagundes David de Oliveira
Other Authors: Ricardo Hiroshi Caldeira Takahashi
Format: Others
Language:English
Published: Universidade Federal de Minas Gerais 2013
Online Access:http://hdl.handle.net/1843/EABA-98VHPQ
Description
Summary:=== This dissertation constructs optimal root-searching and aximumsearching algorithms in a statistical sense and compares the statistically optimal strategies to the already known mini-maximal strategies. In order to construct the so called statistical method, new results in the eld of probability, capable of determining the probability of f(x) = y over a pre-determined set of functions, are presented. === Esta dissertação constrói algoritmos de busca de raiz e de busca de máximos, ótimos em um sentido estatístico, e compara os métodos estatisticamente ótimos com as já conhecidas estratégias mini-maximais. A fim de construir o chamado método estatístico, novos resultados na área de probabilidade, capazes de determinar a probabilidade de f(x) = y sobre um conjunto pré-determinado de funções, são apresentados.