Analysis framework for adaptive spiking neural networks

Spiking Neural Network;Phenomenological models;Bottom-up;Learning;Adaptation;Closed-loop;Parallel;Asynchronous;Simulation

Wang, Felix ; Levinson ; Stephen E.

关键词: Spiking Neural Network;    Phenomenological models;    Bottom-up;    Learning;    Adaptation;    Closed-loop;    Parallel;    Asynchronous;    Simulation;   

Others  :  https://www.ideals.illinois.edu/bitstream/handle/2142/50483/Felix_Wang.pdf?sequence=1&isAllowed=y

美国|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Learning is an inherently closed-loop process that involves the interaction between an intelligent agent and its environment. In the human brain, we assert that the basis for learning is in its ability to represent external stimuli symbolically in an associative memory. Historically, statistical methods such as the hidden Markov model have been used in order to provide the internal symbolic representation to external signals from the environment. This work approaches similar themes by investigating the function of the neocortex, with the ultimate goal of understanding how mental states might arise from spiking activity. Cortical modeling has traditionally focused on the mechanisms and behaviors at the cellular level. However, developments with respect to group or population level phenomena indicate that a shift in focus is necessary to understand how learning and representation of stimuli might occur in the brain. We present a Simulation Tool for Asynchronous Cortical Streams (STACS) for studying spiking neural networks exhibiting adaptation in a closed-loop system.