Bioinspired Learning

Bio-inspired algorithms application por MUTSC/BINL practical project.

In this project we developed a modification for 'Reversing the game of life using genetic algorithms' [1] original implementation. We introduced a regularization term in the fitness function in order to select better initial boards of Conway's game of life based on the concept of 'Garden of Eden'. Our modification allows us to obtain less noisy initial boards that will improve the metrics of the comparision between the 'ground truth' initial board and the predicted board based on the genetic algorith.

We wrote a small paper that can be found An experiment on reversing Game Of Life using Genetic Algorithms

Contact

For further contact you can reach us via e-mail:

Isabel Prieto Payo: [email protected] Víctor Gutiérrez García: [email protected]

Setup

Go to your project parent directory and create and activate your virtual environment.

python3 -m venv env
source env/bin/activate

Once activated, downoload all the required packages.

pip install -r requirements.txt

Now, you will be able to install packages and run Python within the environment without interfering with packages installed globally.

Once you are finished, exit the virtual environment.

deactivate

Developer

You may need to update the requirements file with new needed packages.

pip freeze > requirements.txt

References

[1] Pavel Tyshevskyi. Reversing the game of life using genetic algorithms. https://medium.com/@ptyshevs/rgol-ga-1cafc67db6c7