Runtime phylogenetic analysis enables extreme subsampling for test-based problems

Created by W.Langdon from gp-bibliography.bib Revision:1.8051

@InProceedings{lalejini:2024:GECCOcomp,

• author = "Alexander Lalejini and Marcos Sanson and Jack Garbus and Matthew Andres Moreno and Emily Dolson",
• title = "Runtime phylogenetic analysis enables extreme subsampling for test-based problems",
• booktitle = "Proceedings of the 2024 Genetic and Evolutionary Computation Conference Companion",
• year = "2024",
• editor = "Ting Hu and Aniko Ekart",
• pages = "511--514",
• address = "Melbourne, Australia",
• series = "GECCO '24",
• month = "14-18 " # jul,
• organisation = "SIGEVO",
• publisher = "Association for Computing Machinery",
• publisher_address = "New York, NY, USA",
• keywords = "genetic algorithms, genetic programming, subsampling, selection schemes, lexicase selection, phylogenetic analysis, test-based problems: Poster",
• isbn13 = "979-8-4007-0495-6",
• DOI = "doi:10.1145/3638530.3654208",
• size = "4 pages",
• abstract = "A phylogeny describes a population's evolutionary history. Evolutionary search algorithms can perfectly track the ancestry of candidate solutions, illuminating a population's trajectory through the search space. We introduce phylogeny-informed subsampling, a new class of subsampling methods that exploit runtime phylogenetic analyses for solving test-based problems. Specifically, we assess two phylogeny-informed subsampling methods---individualized random subsampling and ancestor-based subsampling---on ten genetic programming (GP) problems from program synthesis benchmark suites. Overall, we find that phylogeny-informed subsampling methods enable problem-solving success at extreme subsampling levels where other subsampling methods fail. For example, phylogeny-informed subsampling methods more reliably solved program synthesis problems when evaluating just one training case per-individual, per-generation. However, at moderate subsampling levels, phylogeny-informed subsampling generally performed no better than random subsampling on GP problems. Continued refinements of phylogeny-informed subsampling techniques offer a promising new direction for scaling up evolutionary systems to handle problems with many expensive-to-evaluate fitness criteria.",
• notes = "GECCO-2024 GP A Recombination of the 33rd International Conference on Genetic Algorithms (ICGA) and the 29th Annual Genetic Programming Conference (GP)",

}

Genetic Programming entries for Alexander Lalejini Marcos Sanson Jack Garbus Matthew Andres Moreno Emily Dolson

Citations