Modelamiento estocástico de neuronas biológicas usando procesos puntuales
Cargando...
Archivos
Autores
Director
Tipo de contenido
Editor
Document language:
Español
Fecha
Título de la revista
ISSN de la revista
Título del volumen
Documentos PDF
Resumen
En la actualidad están siendo diseñados modelos para redes de neuronas con el objetivo de entender el funcionamiento del cerebro. En la tesis se usa el modelo para una neurona desarrollado por Vidibyda y Arunachalam, esta idea es utilizada en una red de neuronas llamada circuito WTA. Se deducen las propiedades del modelo combinado. Se desarrollan fórmulas para modelos con retroacción instantánea y sin ella. Por último, Se insertan los valores esperados de disparos del circuito en el modelo de Reloj de Arena (Cottrell, 1994) de manera computacional y se comparan los resultados con el análisis hecho por Hesam (2015) para este tipo de circuitos.
Abstract. Models for many neurons connected are required nowadays in order to model the brain. Here, it is used the simple model to one neuron by Vidibyda and Arunachalam and its idea is introduced in the neural network called WTA network and the properties of the new composed model are analyzed. It is assumed both, with instantaneous feedback and without feedback, so equations for each case will be shown. Finally, the firing expect times of the networks are inserted in the hourglass model (made by Cottrell in 1994) computationally and the results are compared with the rhythmic inhibition simulation analysis for neural networks made by Hesham et al.
Abstract. Models for many neurons connected are required nowadays in order to model the brain. Here, it is used the simple model to one neuron by Vidibyda and Arunachalam and its idea is introduced in the neural network called WTA network and the properties of the new composed model are analyzed. It is assumed both, with instantaneous feedback and without feedback, so equations for each case will be shown. Finally, the firing expect times of the networks are inserted in the hourglass model (made by Cottrell in 1994) computationally and the results are compared with the rhythmic inhibition simulation analysis for neural networks made by Hesham et al.