Neural networks and neurocomputational modeling

Toutounji, Hazem; Hert�g, Loreen; Durstewitz, Daniel

Abstract

This chapter reviews methods of neurocomputational modeling, ranging from biophysically detailed single neuron and synapse models to connectionist?style, abstract network formalisms. These methods form an arsenal of mathematical tools that draw on dynamical systems theory, computational theory, nonlinear optimization, probability theory, and statistics. Together, they provide a common language for addressing phenomena at a wide span of biological scales, from molecular mechanisms describing intracellular signal processing to the brain?wide neural activity producing cognition and behavior. They also form the basis for advanced estimation of model parameters and network structure directly from neural recordings. In conclusion, given the commonalities in mathematical approaches addressed through the text, the necessity for an overarching framework to tackle questions in neurocomputational modeling at different levels of biological detail is emphasized.

Citation

Toutounji, H., Hertäg, L., & Durstewitz, D. (2018). Neural networks and neurocomputational modeling. In . E. Wagenmakers (Ed.), Stevens' handbook of experimental psychology and cognitive neuroscience, Vol. V: Methodology. (4th edition). Wiley. https://doi.org/10.1002/9781119170174.epcn517

• Online Publication Date: Mar 23, 2018
• Publication Date: 2018-04
• Deposit Date: Jul 6, 2020
• Publisher: Wiley
• Edition: 4th edition
• Book Title: Stevens' handbook of experimental psychology and cognitive neuroscience, Vol. V: Methodology
• ISBN: 9781119170167
• DOI: https://doi.org/10.1002/9781119170174.epcn517
• Public URL: https://nottingham-repository.worktribe.com/output/4754313
• Publisher URL: https://onlinelibrary.wiley.com/doi/10.1002/9781119170174.epcn517