The final part of the Artificial Intelligence bachelor is a project, with accompanying thesis. Together with two friends, Jaldert Rombouts and Tessa Verhoef, we tried to see the influence of a predator-prey dynamic on speciation by simulating evolution. To assure that speciation would occur from co-evolution the model did not have location. In our simulations we saw that speciation indeed occurs within a wide range of parameters. Our advisor during this project was Bart de Boer.
On the right is an example of the output our software (written in C#) gave. It shows the ‘prey’ on the left (in blue) and the predator on the right (red). The vertical axis is time (every dot is an individual on this timeline), the horizontal position is a certain variable indicating the defence mechanism of the prey. What can be seen is that the defence mechanism drifts away and the predators adapt to this, this is why both lines show the same pattern in general. Half-way thru, however, a speciation can be seen, the blue line diverts into two different species.
The report can be downloaded here (Dutch). This report and the picture in the top are not under the Creative Commons license