GACL - Genetic Algorithms in Conwawy's Life
by David Rodrigues; version 0.9a
Built with Processing
Description
The above applet shows the Genetic Algorithm of a population of Conway's Life Cellular Automata!
The 3 boards are:
- Left: evolution of the present Life Automaton
- Middle: The Initial Chromosome being evolved
- Right: Best Chromosome
The legend corresponds to:
- F: the value of the fitness function of the Automaton being run
- B: the value of the best fitness
- T: the sum of the fitness off the entire population
- g: generation
- c: element of the population being evaluated
Fitness
The fitness function used is the Mutual Information of consecutive states. The fitness function is one of the measures that can be used to identify interesting dynamics in cellular automata (for example)
Population
The population is composed by 20 Life CAs with 192 automata width. They are evolved for 40 timesteps and at the end the value of the Mutual Information of consecutive states is taken as the fitness value.
A one-point crossover is applied and a mutation factor of 1% is applied to the generation+1!