Dynamics of Hierarchical Coupled Neurons

Abstract

It goes without saying that the brain is the most complex system known to humans, and newlineit continues to be at the centre of all fundamental research in the twenty-first century. One of newlinethe most important research endeavours of modern science is the simulation of the human brain newlineand brain functioning. On a global scale, the brain is undoubtedly a weakly connected system; newlinenonetheless, anatomical evidence implies the existence of processing units in a hierarchical newlineorganisation. Massive portions of the brain are involved in the perception of even the most newlinebasic forms of sensory processes. It is also critical to recognise that the neuronal simulation newlineprocess has nothing to do with AI or Strong AI. newlineThe brain can be considered as an oscillatory system that is capable of generating low newlineand high dimensional chaos, which can be measured from the properties of EEG signals. The newlinebrain, being a highly nonlinear system comprising of mass of interactive ensemble of neurons, newlinecomprehending the chaotic dynamics of the brain can be done by observing the attractor newlinedynamics and by measuring the dimension and the amount of unpredictability contained in the newlineEEG signal. newlineThe subject of investigation of the thesis is to comprehend the dynamics of a tiny world newlineof interactive neurons connected in a hierarchical fashion with feedback and feedforward newlineconnections. To accomplish this as one of the objectives of this research, the literature survey newlinedirected us towards a system which is overlooked by many researchers, namely the olfactory newlinesystem. It is known to be simple, stable and presumably less complex in comparison with newlineneocortex and paleocortex and at the same time not compromising on the processing capability newlineof information. The mesoscopic Freeman KIII model which is based on the olfactory system newlinecytoarchitecture, has been found to be one of the best model which could mimic the chaotic newlineactivity of various brain states. In this research, the KIII model is implemented in MATLAB newlinecomprising of KO,KI,KII sets with each layer representing each of