{"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/nsga-net-a-multi-objective-genetic-algorithm","title":"NSGA-Net: Neural Architecture Search using Multi-Objective Genetic Algorithm","arxiv_id":"1810.03522","date":"2018-10-08","proceeding":null,"authors":["Zhichao Lu","Ian Whalen","Vishnu Boddeti","Yashesh Dhebar","Kalyanmoy Deb","Erik Goodman","Wolfgang Banzhaf"],"abstract":"This paper introduces NSGA-Net -- an evolutionary approach for neural\narchitecture search (NAS). NSGA-Net is designed with three goals in mind: (1) a\nprocedure considering multiple and conflicting objectives, (2) an efficient\nprocedure balancing exploration and exploitation of the space of potential\nneural network architectures, and (3) a procedure finding a diverse set of\ntrade-off network architectures achieved in a single run. NSGA-Net is a\npopulation-based search algorithm that explores a space of potential neural\nnetwork architectures in three steps, namely, a population initialization step\nthat is based on prior-knowledge from hand-crafted architectures, an\nexploration step comprising crossover and mutation of architectures, and\nfinally an exploitation step that utilizes the hidden useful knowledge stored\nin the entire history of evaluated neural architectures in the form of a\nBayesian Network. Experimental results suggest that combining the dual\nobjectives of minimizing an error metric and computational complexity, as\nmeasured by FLOPs, allows NSGA-Net to find competitive neural architectures.\nMoreover, NSGA-Net achieves error rate on the CIFAR-10 dataset on par with\nother state-of-the-art NAS methods while using orders of magnitude less\ncomputational resources. These results are encouraging and shows the promise to\nfurther use of EC methods in various deep-learning paradigms.","url_abs":"http://arxiv.org/abs/1810.03522v2","url_pdf":"http://arxiv.org/pdf/1810.03522v2.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":"nsga-net-a-multi-objective-genetic-algorithm","repo_url":"https://github.com/ianwhale/nsga-net","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"nsga-net-a-multi-objective-genetic-algorithm","repo_url":"https://github.com/xuanhungho/nsga","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"efficient-exploration","task_name":"Efficient Exploration"},{"task_slug":"architecture-search","task_name":"Neural Architecture Search"}],"methods":[{"method_slug":"entropy-regularization","method_name":"Entropy Regularization"},{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"neural-architecture-search","method_name":"Neural Architecture Search"},{"method_slug":"ppo","method_name":"PPO"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1810.03522","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}