A golang demo which explores evolutionary algorithms to generate an equation which solves to a given integer.
I originally started this project in 2015 to learn the go programming language and a little about generative AI at the same time. At the time, I followed a tutorial on genetic (evolutionary) algorithms with all code examples written in Java and ported the code to go. I've since revisited the project and refactored the original code to be more go idiomatic. The entire project is simply for fun and education.
This demo is limited to generating equations which solve to relatively small numbers due to the chromosome length and the current algorithm only allowing single digit operands.
The sub-operations in the generative equations solve to
integers. This leads to potential situations where 9 / 2 = 4. If you encounter
a generative equation which is incorrect, this is most likely the cause.
I have begun work porting the original string gene encoding implementation to a byte
encoding implementation. I plan to benchmark the two implemtnations to explore the suspected
performance enhancements of the bytes solution.