{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/data-efficient-neuroevolution-with-kernel","title":"Data-efficient Neuroevolution with Kernel-Based Surrogate Models","arxiv_id":"1804.05364","date":"2018-04-15","proceeding":null,"authors":["Adam Gaier","Alexander Asteroth","Jean-Baptiste Mouret"],"abstract":"Surrogate-assistance approaches have long been used in computationally\nexpensive domains to improve the data-efficiency of optimization algorithms.\nNeuroevolution, however, has so far resisted the application of these\ntechniques because it requires the surrogate model to make fitness predictions\nbased on variable topologies, instead of a vector of parameters. Our main\ninsight is that we can sidestep this problem by using kernel-based surrogate\nmodels, which require only the definition of a distance measure between\nindividuals. Our second insight is that the well-established Neuroevolution of\nAugmenting Topologies (NEAT) algorithm provides a computationally efficient\ndistance measure between dissimilar networks in the form of \"compatibility\ndistance\", initially designed to maintain topological diversity. Combining\nthese two ideas, we introduce a surrogate-assisted neuroevolution algorithm\nthat combines NEAT and a surrogate model built using a compatibility distance\nkernel. We demonstrate the data-efficiency of this new algorithm on the low\ndimensional cart-pole swing-up problem, as well as the higher dimensional\nhalf-cheetah running task. In both tasks the surrogate-assisted variant\nachieves the same or better results with several times fewer function\nevaluations as the original NEAT.","url_abs":"http://arxiv.org/abs/1804.05364v2","url_pdf":"http://arxiv.org/pdf/1804.05364v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"data-efficient-neuroevolution-with-kernel","repo_url":"https://github.com/agaier/matNEAT","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"diversity","task_name":"Diversity"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}