Training an agent on brainwaves: using brain signals as feedback for reinforcement learning

"This thesis replicates and proposes an alternative method to train reinforcement learning algorithms with ErrP signals, captured through EEG, and validate the effectiveness of its use in a prototype application."

Guardado en:
Detalles Bibliográficos
Autores principales: Moreno, Juan, Bartolomé, Francisco, Navas, Natalia, Vitali, José
Otros Autores: Ramele, Rodrigo
Formato: Proyecto final de Grado
Lenguaje:Español
Publicado: 2019
Materias:
Acceso en línea:http://ri.itba.edu.ar/handle/123456789/1785
Aporte de:
Descripción
Sumario:"This thesis replicates and proposes an alternative method to train reinforcement learning algorithms with ErrP signals, captured through EEG, and validate the effectiveness of its use in a prototype application."