Neuro-Cellular Automata: Difference between revisions
From Santa Fe Institute Events Wiki
No edit summary |
No edit summary |
||
| (5 intermediate revisions by the same user not shown) | |||
| Line 1: | Line 1: | ||
{{CSSS 2006--Santa Fe}} | {{CSSS 2006--Santa Fe}} | ||
''' | '''Neuro-Cellular Automata''' | ||
A new type of Cellular Automata is presented and studied: the Neuro-Cellular Automata. The standard way of representing the rules of an automaton is replaced by an Artificial Neural Network. The motivation of this change is two-fold: to have a more biologically inspired automata and to explore the connections between Cellular Automata, Neural Networks, Evolution and Complexity. The process of the development of this kind of automata is also described, as well as some considerations about important aspects of the work. We use simple one-dimensional Cellular Automata, in order to better understand the complexity of the different components of the Neuro-Cellular Automata and to evolve it, because evolution is required to design these type of automata. A study regarding ways to measure complexity is also performed, since these measures play an important role in evaluating the population of automata in the evolutionary process. | |||
'' | ''Participants'': [[Jorge Tavares]], [[Cornelia Kreutzer]], [[Anna Fedor]] | ||
''Paper'': [[Media:sf_csss06_tavares_et_al.pdf | Neuro-Cellular Automata: Connecting Cellular Automata, Neural Networks and Evolution ]] | |||
Back to [[Group Pages]] | Back to [[Group Pages]] | ||
Latest revision as of 19:13, 9 July 2007
| CSSS 2006 Santa Fe |
Neuro-Cellular Automata
A new type of Cellular Automata is presented and studied: the Neuro-Cellular Automata. The standard way of representing the rules of an automaton is replaced by an Artificial Neural Network. The motivation of this change is two-fold: to have a more biologically inspired automata and to explore the connections between Cellular Automata, Neural Networks, Evolution and Complexity. The process of the development of this kind of automata is also described, as well as some considerations about important aspects of the work. We use simple one-dimensional Cellular Automata, in order to better understand the complexity of the different components of the Neuro-Cellular Automata and to evolve it, because evolution is required to design these type of automata. A study regarding ways to measure complexity is also performed, since these measures play an important role in evaluating the population of automata in the evolutionary process.
Participants: Jorge Tavares, Cornelia Kreutzer, Anna Fedor
Paper: Neuro-Cellular Automata: Connecting Cellular Automata, Neural Networks and Evolution
Back to Group Pages