2025-10-092025-10-092015-12-03COELHO, Pâmela da Silva. Previsão do preço médio de ação preferencial: estudo em empresa atuante na BM&F BOVESPA mediante rede neural artificial Polinomial Group Method of Data Handling (GMDH). 2015. Trabalho de Conclusão de Curso (Graduação em Engenharia de Produção) – Universidade do Estado do Pará, Belém, 2015.https://bibliotecadigitaldetcc.uepa.br/handle/riuepa/686The theme of the paper performed was the forecast of average price behavior of a preferential share of company listed in the Stock Market, Goods and Futures of São Paulo. The objective was the study of the conduct and to make the forecast of average price behavior of share with the application in the polynomial neural network Group Method of Data Handling (GMDH). For this reason there were collected 300 average price values available in the BM&F Bovespa site, organized in two volumes of work (training and net test). The GMDH network was applied to the set of information using the software Matlab R2011b. Finally, obtaining the following values of mean squared error, root mean squared error, error mean, standard deviation and mean absolute percentage error for the respective test data: 0,06681345; 0,25848; -0,086406; 0,245704; 2,1327%.Língua PortuguesaAbertoLogistical componentsQuality of servicesReturnsTrabalho de Conclusão de Curso (TCC)