Diseño de un recurso educativo orientado a docentes e investigadores en formación en ciencias naturales e ingeniería, para el estudio del funcionamiento computacional de una red neuronal

Abstract

Science and technology have been of great influence in the society of the 21st century and little by little teacher training has become a part. The existing material, aimed at teachers and researchers is scarce and limited to manuals, which does not provide an academic training that encourages the research development of teachers and researchers in the area of ​​science and engineering. In the line of scientific and technological literacy that leads the FISINFOR group, we see the need to generate new resources in front of the computational area of ​​neural networks, which favor scientific-technological development in that area. Taking into account the above, this degree work contains an introduction on basic concepts of the brain and the physiology of the neuron with its general functioning and a historical context of the neural network. To make way for the biological and artificial system by describing the artificial neuron model and the activation function. In addition, the architecture of a biological neuron is described, and artificially, finally, it is answered what is the need that replaces the development of the neural network in which concepts of topology, types of architecture and learning will be exposed, finally The backpropagation algorithm is described from its structure and learning.