{"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/accelerating-the-evolution-of-convolutional","title":"Accelerating the Evolution of Convolutional Neural Networks with Node-Level Mutations and Epigenetic Weight Initialization","arxiv_id":"1811.08286","date":"2018-11-17","proceeding":null,"authors":["Travis Desell"],"abstract":"This paper examines three generic strategies for improving the performance of\nneuro-evolution techniques aimed at evolving convolutional neural networks\n(CNNs). These were implemented as part of the Evolutionary eXploration of\nAugmenting Convolutional Topologies (EXACT) algorithm. EXACT evolves arbitrary\nconvolutional neural networks (CNNs) with goals of better discovering and\nunderstanding new effective architectures of CNNs for machine learning tasks\nand to potentially automate the process of network design and selection. The\nstrategies examined are node-level mutation operations, epigenetic weight\ninitialization and pooling connections. Results were gathered over the period\nof a month using a volunteer computing project, where over 225,000 CNNs were\ntrained and evaluated across 16 different EXACT searches. The node mutation\noperations where shown to dramatically improve evolution rates over traditional\nedge mutation operations (as used by the NEAT algorithm), and epigenetic weight\ninitialization was shown to further increase the accuracy and generalizability\nof the trained CNNs. As a negative but interesting result, allowing for pooling\nconnections was shown to degrade the evolution progress. The best trained CNNs\nreached 99.46% accuracy on the MNIST test data in under 13,500 CNN evaluations\n-- accuracy comparable with some of the best human designed CNNs.","url_abs":"http://arxiv.org/abs/1811.08286v1","url_pdf":"http://arxiv.org/pdf/1811.08286v1.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":"accelerating-the-evolution-of-convolutional","repo_url":"https://github.com/travisdesell/exact","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}