Data mining interesting "patterns"

[stdm]

Dear community,

from time to time I enjoy watching the evolution of a quite small (e.g., 23x79 cells) game of life board that has been randomly initialized (50% chance for each cell in the first generation to be alive). I like to see the ermerging and collapse of certain structures. As I have read, a lot of people are into studying (isolating / designing / ...) such patterns, but as far as I understood it, most of this is done manually.

Having some interest in data mining / machine learning / artificial intelligence, I was wondering if you know of any work done in analyzing data that represents the evolution in time of such a delimited life board?

E.g., such a data set of a "game of life log file" could look like
[vector containing the state of any cell as 0 or 1 in generation 1]
[vector containing the state of any cell as 0 or 1 in generation 2]
...
[vector containing the state of any cell as 0 or 1 in generation T]
-> This would yield a data set with 23x79 columns and T rows.

This could be subject to any algorithm that finds repeating or interesting patterns (still, oscillators, spaceships?)...

Do you know of any such work?
Looking into this matter myself brought forth just this article: http://www.researchgate.net/publication ... 7a01e6.pdf

Best regards,
Thilo