Search Results

Cooperation-induced Criticality in Neural Networks

Description: The human brain is considered to be the most complex and powerful information-processing device in the known universe. The fundamental concepts behind the physics of complex systems motivate scientists to investigate the human brain as a collective property emerging from the interaction of thousand agents. In this dissertation, I investigate the emergence of cooperation-induced properties in a system of interacting units. I demonstrate that the neural network of my research generates a series of properties such as avalanche distribution in size and duration coinciding with the experimental results on neural networks both in vivo and in vitro. Focusing attention on temporal complexity and fractal index of the system, I discuss how to define an order parameter and phase transition. Criticality is assumed to correspond to the emergence of temporal complexity, interpreted as a manifestation of non-Poisson renewal dynamics. In addition, I study the transmission of information between two networks to confirm the criticality and discuss how the network topology changes over time in the light of Hebbian learning.
Date: August 2013
Creator: Zare, Marzieh
Partner: UNT Libraries

Complexity matching in neural networks

Description: This article adopts the complexity matching principle that the maximal efficiency of communication between two complex networks is realized when both of them are at criticality, and uses this principle to establish the value of the neuronal interaction strength at which criticality occurs.
Date: January 9, 2015
Creator: Mafahim, Javad Usefie; Lambert, David; Zare, Marzieh & Grigolini, Paolo
Partner: UNT College of Arts and Sciences