Modeling challenge

A hands-on aspect of the proposed seminar will be the modeling challenge. The idea is to select a handful of concrete problems that lend themselves to computational, and, ideally, embodied computational, modeling and do actual work on these problems at the seminar. However, it would be naive to assume that real progress can be made during the allocated time (3 afternoons essentially). Therefore, we are planning to set the ground for this challenge prior to the seminar and to continue working after the seminar.

Datasets

A number of datasets will be made available by the participants. You can contribute your own - the template to desribe it is roughly:

Title; Short description; [Picture / Schematics / Setup]; Data description (E.g. a .cls file with every column representing…. / a spike train / ...); Questions, hypotheses; What can we learn from (embodied) computational model; Requirements on the model (E.g. a humanoid robot with sensitive skin / a foetus model and a spiking neural network / ...); Existing publications on this; Link to videos / dataset and additional materials; Contact;

Restrictions on who can access the data;

Modeling break-out groups

So far, the following groups have been formed (7.10.2018):

• Learning peripersonal space representations (moderator Jean-Paul Noel)
• Rubber hand illusion models (moderator Pablo Lanillos)
• Sensorimotor self and its development (moderators Matej Hoffmann / Matthew Longo / Gregor Schöner)