Information Coding in Neural Models of Spiking Elements with External Forcing
نویسنده
چکیده
Different encoding schemes are applied in neural network modelling at the level of a single neuron. (1) Fine temporal coding. The fine temporal structure of neuronal spiking is used as a basis for coding and information processing (Mainen & Sejnowski, 1995). (2) Rate coding. A single neural spiking rate is used as a code. This encoding scheme is rough because the temporal pattern of spiking is neglected and small variations of spike times do not change the rate code (Shadlen & Newsome, 1998). (3) Phase-frequency coding. There is experimental evidence that neurons have frequency preference due to a resonance between the input signal and internal oscillations of the neuronal membrane potential (Hutcheon & Yarom, 2000). This approach to neuronal coding seems to be very promising both from neuroscientific and mathematical/computational points of views. Neural information coding can be implemented not only at the level of individual neurons but also at the level of activity of neural assemblies. The ideas about information coding by neural populations are similar to those of individual neurons: (1) Spatio-temporal coding. The characteristics of spatio-temporal patterns of neural activity determine the code. Time and space variation of dynamics of neural activity defines functional correlates of information processing. For example, the synchrony of spike trains is related to learning and memory. (2) Population rate coding. The rate/activity at time t is determined as the average number of firing neurons in a given population per a time unit (or as the average membrane potential of neurons at the moment t). (3) Phase-frequency coding in oscillatory neural networks. The phases of oscillations and the natural frequencies of oscillators are characteristic variables of this approach to neuronal coding. The above-mentioned coding schemes can be realized in the frames of the following theoretical construction. A coding system is implemented by a network (or a single neuron) of dynamically interacting elements supplied by a set of inputs (which deliver the information about a stimulus to the network) and a set of outputs (which transmit the results of coding to other information processing systems). Depending on the type of coding, the elements of the network simulate in different detail the functioning of individual neurons, neural populations (excitatory and inhibitory), or neural structures. As the input signals, both constant and changing continuous signals can be used, as well as stochastic or deterministic sequences of spikes. The signals at different inputs can be identical to …
منابع مشابه
Spiking AGREL
Spiking neural networks are characterised by the spiking neuron models they use and how these spiking neurons process information communicated through spikes – the neural code. We demonstrate a plausible spiking neural network based on Spike Response Models and predictive spike-coding. When combined with a plausible reinforcement learning strategy – Attention Gated REinforcement Learning (AGREL...
متن کاملImplementing Signature Neural Networks with Spiking Neurons
Spiking Neural Networks constitute the most promising approach to develop realistic Artificial Neural Networks (ANNs). Unlike traditional firing rate-based paradigms, information coding in spiking models is based on the precise timing of individual spikes. It has been demonstrated that spiking ANNs can be successfully and efficiently applied to multiple realistic problems solvable with traditio...
متن کاملSupervised and unsupervised weight and delay adaptation learning in temporal coding spiking neural networks
Artificial neural networks are learning paradigms which mimic the biological neu ral system. The temporal coding Spiking Neural Network, a relatively new artifi cial neural network paradigm, is considered to be computationally more powerful than the conventional neural network. Research on the network of spiking neurons is an emerging field and has potential for wider investigation. This rese...
متن کاملA Spike-Timing-Based Integrated Model for Pattern Recognition
During the past few decades, remarkable progress has been made in solving pattern recognition problems using networks of spiking neurons. However, the issue of pattern recognition involving computational process from sensory encoding to synaptic learning remains underexplored, as most existing models or algorithms target only part of the computational process. Furthermore, many learning algorit...
متن کاملA memristive spiking neuron with firing rate coding
Perception, decisions, and sensations are all encoded into trains of action potentials in the brain. The relation between stimulus strength and all-or-nothing spiking of neurons is widely believed to be the basis of this coding. This initiated the development of spiking neuron models; one of today's most powerful conceptual tool for the analysis and emulation of neural dynamics. The success of ...
متن کامل