CNS*2003
    


CNS*2003 Workshops

University Miquel Hernández, 8-9 July 2003

Unifying Neural Coding, Computation and Dynamics (Tutorial)

  • C. H. Anderson (Washington University, St. Louis, USA)
  • C. Eliasmith (University of Waterloo, Waterloo, Canada)

For many years, researchers have used well-established quantitative methods for trying to better understand neural systems. These methods are derived from information theory, control theory, signal processing, dynamic systems theory, probability theory, statistical mechanics, and so on. However, these various methods have yet to be integrated into a unified framework for characterizing neural systems in general.

In this workshop, we present our attempt at developing such a framework, which has been presented and developed in detail in Eliasmith & Anderson (2003) Neural Engineering: Computation, Representation and Dynamics in Neurobiological Systems (MIT Press). In the workshop we will review the main principles of the framework, and describe a methodology for applying these principles to generate large-scale, spiking models of various nerual systems. These principles address central issues in population coding, temporal coding, neural dynamics, data analysis, and experimental design.

We will present examples of diverse neural systems, including perceptual (vestibular), cognitive (working memory), and motor (eye position control) systems. The examples demonstrate how the framework: 1) helps make specific predictions regarding these systems; 2) can be used to generate models of nonlinear transformations; 3) supports linear and nonlinear control structures; 4) is consistent with various neuron models (from LIF to conductance-based models); and 5) supports concurrent modeling of a system at many levels of description.

In addition, we will give a brief introduction to the software package NESims that has been developed in conjunction with this framework.

See here for more information.

The organizers will present the overview and examples described above. We invite the submission of talks based on analyses or models related to this framework. The deadline for submissions is June 15, 2003.