Análisis comparativo de algoritmos neuro computacionales biológicos para proceso cognitivo de la memoria y su aplicación en el robot Nao
This article presents a detailed neuro-computational algorithms study, the first based on reinforcement learning called Q-learning and the second based on information or updated algorithm called A* Star, and then compare them in principle from its theoretical concepts through its mathematical form a...
| Autor Principal: | Gómez Vega, Andrés Santiago |
|---|---|
| Otros Autores: | Guerrero Rivera, Omar Fernando |
| Formato: | bachelorThesis |
| Idioma: | spa |
| Publicado: |
2016
|
| Materias: | |
| Acceso en línea: |
http://dspace.ups.edu.ec/handle/123456789/13083 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: |
This article presents a detailed neuro-computational algorithms study, the first based on reinforcement learning called Q-learning and the second based on information or updated algorithm called A* Star, and then compare them in principle from its theoretical concepts through its mathematical form and then with the prior knowledge of its structure, implement them in a humanoid robot NAO. This implementation is carried out by programming software called choreographe, which is understandable to the robot named Naoqi and allows the programming language Python. First the algorithms used individually to verify their behavior in the humanoid robot NAO and identify efficiency in the interaction in the real world, for which an environment is created in the form of maze. With both algorithms, A* and Q-Learning, compares two parameters: runtime and number of iterations to complete learning. |
|---|