La titularidad de los derechos patrimoniales de esta obra pertenece a las instituciones editoras. Su uso se rige por una licencia Creative Commons BY-NC-ND 4.0 Internacional, https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode.es, para un uso diferente consultar al responsable jurídico del repositorio por medio del correo electrónico editor_general@rmac-mx.org.
Ver términos de la licencia
Burgos, José E.
Tipo
Artículo de Investigación
Área del conocimiento
Medicina y Ciencias de la Salud
Título
A biobehavioral approach to an aspect of social behavior
Fecha
2011-01-26
Resumen
The present paper describes a biobehavioral approach to an aspect of social behavior, namely, learning to respond to the ongoing behavior of another individual. The approach was implemented through computer simulations that involved a combination of a neurocomputational model, a network model, a neurodevelopmental model, and a genetic algorithm. In Phase 1 of the core simulation, ten 50-generation lineages evolved under a Pavlovian procedure with one conditional stimulus (CS1). Each lineage had its own random founder population of 100 genotypes. In Phase 2, ten genotypes were randomly chosen from the last generation of each lineage, to form the founder population for a new lineage. In each generation of this lineage, individuals were randomly selected with a small probability to function as "senders". Senders were first trained under the same arrangement as their ancestors. Then, they were given 100 maintenance trials under the same arrangement, during which their output activations in the presence of the cs1 served as a cs2 for the rest of the population, which functioned as "receivers". All individuals were selected for high conditional responding to their respective cs. Results showed that selection for responding to the behavior of another network reduced population genetic and phenetic variation and increased mean population fitness across generations.
Tema
Social Behavior; Biobehavioral Approach; Computer Simulations; Pavlovian Conditioning; Evolution; Artificial Neural Networks; Genetic Algorithms; Conducta Social; Aproximación Bioconductual; Simulaciones Por Computadora; Condicionamiento Pavloviano; Evolución; Redes Neurales Artificiales; Algoritmos Genéticos
Idioma
eng
ISSN
ISSN: 0185-4534; ISSN electrónico: 2007-0802
