Tag-based Module Regulation for Genetic Programming

This Hot-off-the-Press paper summarizes our recently published work, “Tag-based regulation of modules in genetic programming improves context-dependent problem solving,” published in Genetic Programming and Evolvable Machines. We introduce and experimentally demonstrate tag-based genetic regulation, a genetic programming (GP) technique that allows programs to dynamically adjust which code modules to express. Tags are evolvable labels that provide a flexible naming scheme for referencing code modules. Tag-based regulation extends tag-based naming schemes to allow programs to “promote” and “repress” code modules to alter module execution patterns. We find that tag-based regulation improves problem-solving success on problems where programs must adjust how they respond to current inputs based on prior inputs; indeed, some of these problems could not be solved until regulation was added. We also identify scenarios where the correct response to an input does not change over time, rendering tag-based regulation an unnecessary functionality that can sometimes impede evolution. Broadly, tag-based regulation adds to our repertoire of techniques for evolving more dynamic computer programs and can easily be incorporated into existing tag-enabled GP systems.

⎘ copy to clipboard

⎘ copy to clipboard

Alexander Lalejini, Matthew Andres Moreno, and Charles Ofria. 2022. Tag-based Module Regulation for Genetic Programming. In Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO ‘22). Association for Computing Machinery, New York, NY, USA, 25–26. https://doi.org/10.1145/3520304.3534060

• primary publication