Brain Architecture: DAC Grid Cells and Cerebellum

Grid cells in layer II medial entorhinal cortex of the mammalian brain fire periodically tiling the explored environment in a striking triangular tessellation pattern [Hafting et al. 2005]. It has been shown that attractor mechanisms govern the network dynamics of such representational system [Yoon et al. 2013]. A computational model of grid cells based on low-dimensional attractor dynamics has been presented [Guanella et al. 2006]. During this tutorial we will review and implement such model using a simulated agent. We will use Numpy and Matplotlib python libraries for simulation and data visualisation.

During this tutorial we will cover the basics on cerebellar learning focussing on the paradigms of eye blink conditioning and avoidance learning. We' ll analyze the main  behavioral aspects related to these learning paradigms, the underlying cerebellar anatomy and physiology and its functions. We'll then explore how the cerebellar microcircuit can be algorithmically modeled and implemented in mobile robots that are able to adapt their motor control skills by interacting with the world. Finally, if time allows, a brief practical session will be provided on Python / IPython Notebook.

The tutorial on the cerebellum will continue on Thursday 3rd of September with Ivan Herreros.