Wolfram Computation Meets Knowledge

Wolfram Summer School


Atilim Gunes Baydin

Summer School

Class of 2005


Atilim Gunes Baydin graduated from Middle East Technical University in Turkey, with a Bachelor of Science degree in engineering. Just after the NKS Summer School, he will start studying in the master’s program called “Complex Adaptive Systems” at the Chalmers University of Technology and Göteborg University in Sweden. The program involves mainly the theory of complex systems, supported by topics like information theory, computational biology, and artificial intelligence.

He is an experienced programmer with a profound interest in electronics and robotics. He is particularly optimistic that science, possibly in the near future, will become unified–that the many diverse fields in contemporary science will ultimately realize that they just involve specialized versions of more fundamental phenomena. Thinking that such a unity is inevitable, the paradigm shift Dr. Wolfram wants to achieve has the potential to kick-start science down such a path.

His main interests: emergent behavior and self organization, the nature of consciousness and intelligence, mechanisms giving rise to intelligence and machine learning, adaptive and autonomous system design, and simulation of complex systems and chaos.

His other personal interests include history (Hellenistic and Roman times more than others), linguistics (Latin and the Indo-European/Nostratic studies), philosophy, and art.

Project: Systematic Leaf Shape Modeling

This project involves 2D substitution systems that simulate plant growth and the shapes of plant leaves. The project started with the aim of constructing an automated system that, given an image of a real plant leaf, will find the parameters of a substitution model at hand or possibly produce a new model each time–using L Systems–growing into a similar shape. The first idea was to employ genetic algorithms (GA) in matching. At the current stage, the working code of matching works in a way similar to exhaustive search with quantized real-valued parameters and constructing inverse mappings of image space to parameter space.

A new substitution model (an extension to the model constructed for the same purpose in the NKS Book, Chapter 8) is found to be able to produce a great variety of leaf shapes. An additional experiment with GA applied on the proposed model has shown good initial performance.

The proposed new substitution model is capable of reproducing a very high portion of the plant leaves found in the nature.

Favorite Four-Color, Nearest-Neighbor, Totalistic Rule

Rule chosen: 930704

My favorite four-color, nearest-neighbor, totalistic CA rule turns out to be rule 930704–given the random seed of Mathematica and the initial state of my brain, at the time I started my search.