Caracterización del Fenómeno de Aprendizaje en Memorias de Largo Plazo para Arquitecturas Cognitivas de Robots Autónomos

Since the beginnings of Artificial Intelligence, cognitive architectures have been an area of ​​research interest since the 1950s, with the aim of creating computer programs that reason about problems across various domains, develop knowledge, adapt to new situations, and reflect on themselves.

Cognitive abilities such as memory, attention, language, perception, problem-solving, or intelligence and planning involve sophisticated and unique brain functions.

Currently, cognitive robotics is focused on the design of artificial agents capable of performing cognitive tasks autonomously, as in humans. A central issue for this task is studying the processes by which these agents learn.

Learning, as such, is one of the most important characteristics to be improved in the coming years if we want real progress that will allow us to have autonomous agents that will serve in many specialized tasks in the future. Learning is directly related to long-term memory capacity.

This work seeks to provide a significant contribution to current techniques for data management in long-term memories and decision-making, for an improvement in the cognitive processes of autonomous robots

A joint effort is proposed between TEC Costa Rica and the Integrated Engineering Group (GII) of the Universidade da Coruña, Spain, which can be of great benefit to the institution

Desarrollar y validar una representación del conocimiento (o nugget de conocimiento = NC) y algoritmos para la evaluación de características de
interés en Memorias de Largo Plazo para Arquitecturas Cognitivas de
Robots Autónomos.

Teoría y Metodologías en Computación: inteligencia artificial, aprendizaje automático, robótica