{"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/automated-design-using-neural-networks-and","title":"Automated Design using Neural Networks and Gradient Descent","arxiv_id":"1710.10352","date":"2017-10-27","proceeding":"ICLR 2018 1","authors":["Oliver Hennigh"],"abstract":"We propose a novel method that makes use of deep neural networks and gradient\ndecent to perform automated design on complex real world engineering tasks. Our\napproach works by training a neural network to mimic the fitness function of a\ndesign optimization task and then, using the differential nature of the neural\nnetwork, perform gradient decent to maximize the fitness. We demonstrate this\nmethods effectiveness by designing an optimized heat sink and both 2D and 3D\nairfoils that maximize the lift drag ratio under steady state flow conditions.\nWe highlight that our method has two distinct benefits over other automated\ndesign approaches. First, evaluating the neural networks prediction of fitness\ncan be orders of magnitude faster then simulating the system of interest.\nSecond, using gradient decent allows the design space to be searched much more\nefficiently then other gradient free methods. These two strengths work together\nto overcome some of the current shortcomings of automated design.","url_abs":"http://arxiv.org/abs/1710.10352v1","url_pdf":"http://arxiv.org/pdf/1710.10352v1.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":"automated-design-using-neural-networks-and","repo_url":"https://github.com/loliverhennigh/Steady-State-Flow-With-Neural-Nets","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}