Talk by Dr. Martin Pyka
Mercator Research Group "Structure of Memory", Faculty of Psychology, Ruhr-University Bochum, Germany
- 05 Jul 2013 13:30
- 05 Jul 2013 15:00
- INM-6, Bldg. 15.22, Seminar Room 3009, 1. OG
Linking developmental approaches with large-scale perspectives on the human brain
Neuroscientific research is driven by a multitude of perspectives on the human brain that determine study designs and the conclusions we draw from them. The development of a computational model that unifies these perspectives requires an understanding for the various scales in which the brain is analyzed and the corresponding categories in which networks are assessed. Moreover, for the design of artificial neural networks, it remains a challenging task to find the right trade-off between biological correctness and effective abstraction.In this talk, I first give a short overview of some of my studies that I conducted using neuroimaging-techniques (e.g. fMRI) and computational methods to outline large- and small-scale perspectives on the brain. Subsequently, I introduce an evolutionary developmental perspective on neural networks. Approaches and techniques inspired by gene-regulatory networks and self-organizing principles will be presented that can be used to encode functional principles rather than specific neuron-to-neuron connections. At the end, arguments are discussed why developmental approaches for designing artificial neural networks might prove valuable to understand natural neural networks on various scales.