My Problems..

Eytan Ruppin, CS and Medicine, TAU

Abstract:

I imagine that many of you know about my work as little as i know about yours. Its a natural situation, perhaps, in our ``specialized world'' -- certainly sufficient, but perhaps not necessary.. In this talk I would like to try and remedy this, by telling you about some of the problems that interest my students and myself, and about our attempts to address them. The presentation will not be technical (unless you'll ask for, of course, the floor is open) and will be more a conceptual overview, to give you the ``flavor'' of our stuff.

My research in the last years has focused on neurally-driven Evolutionary Autonomous Agents (EAAs). Basically, we address two questions: 1. Can we find ways for developing ``smart'' EAAs? 2. Can we understand whats going on in the controlling networks of these evolved agents? After a brief introduction to the field and ``whats it all about'', I shall aim to review a few of our studies and the (insufficient) answers they give to these questions. I will strive to emphasize the aspects posed by these questions that may be of a challange to computer scientists ``ba asher hem''. I look forward for your feedback, during and after my talk.

The results reviewed are joint work with Isaac Meilijson, Ranit Aharonov, Tuvik Beker, Shlomy Boshy, Zohar Ganon, Yael Niv and Lior Segev.