Análise de um modelo probabilístico de decisão utilizando uma rede de neurônios recorrentes

• Ano de defesa: 2011
• Autor(a) principal: Fernandes, Daniel Abrunhosa
• Orientador(a): Porto, Fabio Andre Machado
• Banca de defesa: Zana, Yossi , Blanco, Pablo Javier , Tort, Adriano Bretanha Lopes
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Laboratório de Computação Científica
• Programa de Pós-Graduação: Programa de Pós-Graduação em Modelagem Computacional
• Departamento: Serviço de Análise e Apoio a Formação de Recursos Humanos
• País: Brasil
• Palavras-chave em Português: Neurociências - Simulação por computador; Processo decisório
• Palavras-chave em Inglês: Decision making; Neurosciences
• Área do conhecimento CNPq: CNPQ::CIENCIAS DA SAUDE::MEDICINA::CLINICA MEDICA::NEUROLOGIA
• Link de acesso: https://tede.lncc.br/handle/tede/147

The area of computational neuroscience has advanced considerably based on the simulation of neuronal networks, formulated from experimental data, and providing findings on the functioning of certain brain areas. This work contributes to this area by studying the process of decision making and its storage in working memory. This work analyzes the results of computer simulations of empirical experiments performed in monkeys. In these experiments, a monkey fixed his attention to a screen where a set of points moved to one of two directions (left or right). From this observation, the monkey decided whether the set of points was heading to the right or left. Based on this experiment, Wang developed a mathematical and computational model that simulates the decision process using a neural network (Wang, 2002). We studied deeper the model proposed by Wang (Wang, 2002) studying factors related to the process of decision making, as well as the behavior of working memory. We implemented some modifications of Wang s model, changing some parameters used in the models. The results suggest that the decrease of the period of stimulus presentation increases the indecision in the model, the insertion of a delay period prior to stimulus presentation has an effect on some aspects of the decision-making process and the model used can not replicate the forgetting of an information in working memory.