Carlos is an Biology undergraduate student in the Universidad Nacional Autonoma de Mexico with interests in systems biology, complexity and ants. His undergraduate dissertation is about the effect of auto-regulation in small genetic regulatory network models, focusing on their stability. When not working on his dissertation, he takes care of a few ant colonies, some of which he started from a single queen and now are a few years old, and, using nature as inspiration, creates jewelry.
Computational Essay: Why are bugs so strong?
Project: Evolving randomly generated programs
Develop a method to create randomly generated programs and to evolve them to do interesting things.
Summary of Results
First, I developed a method to create and mutate random programs. Some of the functions I employed are Reverse, Sort, Accumulate, MinMax, Union, Join and Riffle. After that, I created a fitness function that measures how well a program creates a list of increasing integers. Then I implemented a genetic algorithm to evolve the programs and analyzed the results.
- Refine and increase the set of functions used in random programs.
- Use different sets of functions that work with different objects.
- Change the goal (fitness function) so that the programs evolve to do other interesting things with pseudorandom sequences, the Fibonacci sequence, prime numbers, etc.