Weakly connected neural networks (Q1366659)

This book is devoted to an analysis of general weakly connected neural networks (WCNNs) that can be written in the form \(\dot x_i= f_i(x_i)+ \varepsilon g_i(x_1,\dots, x_n,\varepsilon)\), \(\varepsilon\ll 1\), where each \(x_i\in \mathbb{R}^m\) is a vector that summarizes all physiological attributes of the \(i\)th neuron, \(n\) is the number of neurons, \(f_i\) describes the dynamics of the \(i\)th neuron, \(g_i\) describes the interactions between neurons, and \(\varepsilon\) indicates the strength of connections between neurons. In Chapter 1, the definitions and explanations of basic neurophysiological terminology are given. Several types of mathematical models in the neurosciences are discussed. Bifurcations play an important role in neuron dynamics. This aspect is analysed in Chapter 2. When a neural network's activity is near a bifurcation point, it is sensitive to even small external perturbations. Future behaviour of such networks can be determined by the local events at the bifurcation. Such neural network type -- the nonhyperbolic neural network -- is the topic of discussion of Chapter 3. Theoretical studies of the human brain require mathematical models and it is essential to study families of models instead of one model. To carry out this task, the authors develop an approach that reduces a family of models to a canonical model that has, in some sense, a universal character. This is the main topic of discussion of Chapter 4. Some canonical models for multiple bifurcations in WCNNs are derived in Chapter 5. In Chapter 6, the local dynamics of singularly perturbed WCNNs are studied. While previous chapters discussed the dynamics of WCNNs governed by a system of ordinary differential equations, Chapter 7 considers weakly connected networks of difference equations or mappings. Canonical models for some global bifurcations and for weakly connected oscillatory networks are derived in Chapters 8 and 9, respectively. Canonical models describing oscillatory phenomena are also studied in Chapters 10 and 12 while canonical models for multiple cusp bifurcations are discussed in Chapter 11. Synaptic organizations and dynamical properties of weakly connected neural oscillators are studied in chapter 13. The book ends with a list of useful references. This book is useful to researchers and graduate students of mathematical neurosciences.