{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T15:14:53Z","timestamp":1780586093813,"version":"3.54.1"},"reference-count":86,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","license":[{"start":{"date-parts":[[2021,1,1]],"date-time":"2021-01-01T00:00:00Z","timestamp":1609459200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/legalcode"}],"funder":[{"DOI":"10.13039\/501100004663","name":"Ministry of Science and Technology (MOST), Taiwan","doi-asserted-by":"publisher","award":["109-2221-E-009-116-MY3"],"award-info":[{"award-number":["109-2221-E-009-116-MY3"]}],"id":[{"id":"10.13039\/501100004663","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Access"],"published-print":{"date-parts":[[2021]]},"DOI":"10.1109\/access.2021.3090918","type":"journal-article","created":{"date-parts":[[2021,6,21]],"date-time":"2021-06-21T20:05:51Z","timestamp":1624305951000},"page":"89602-89618","source":"Crossref","is-referenced-by-count":45,"title":["Gradient Descent Effects on Differential Neural Architecture Search: A Survey"],"prefix":"10.1109","volume":"9","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-6710-4846","authenticated-orcid":false,"given":"Santanu","family":"Santra","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jun-Wei","family":"Hsieh","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chi-Fang","family":"Lin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref73","article-title":"Probabilistic neural architecture search","author":"casale","year":"2019","journal-title":"arXiv 1902 05116"},{"key":"ref72","article-title":"DOTS: Decoupling operation and topology in differentiable architecture search","author":"gu","year":"2020","journal-title":"arXiv 2010 00969"},{"key":"ref71","article-title":"StacNAS: Towards stable and consistent optimization for differentiable Neural Architecture Search","volume":"abs 1909 11926","author":"li","year":"2019","journal-title":"CoRR"},{"key":"ref70","first-page":"1","article-title":"The concrete distribution: A continuous relaxation of discrete random variables","author":"maddison","year":"2017","journal-title":"Proc 5th Int Conf Learn Represent (ICLR)"},{"key":"ref76","article-title":"DARTS-: Robustly stepping out of performance collapse without indicators","author":"chu","year":"2021","journal-title":"Proc Int Conf Learn Represent"},{"key":"ref77","article-title":"DRNAS: Dirichlet neural architecture search","author":"chen","year":"2021","journal-title":"Proc Int Conf Learn Represent"},{"key":"ref74","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00783"},{"key":"ref39","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01246-5_2"},{"key":"ref75","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2020\/322"},{"key":"ref38","first-page":"550","article-title":"Understanding and simplifying one-shot architecture search","author":"bender","year":"2018","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1016\/B978-0-12-815480-9.00015-3"},{"key":"ref79","article-title":"Towards automated deep learning: Efficient joint neural architecture and hyperparameter search","author":"zela","year":"2018","journal-title":"arXiv 1807 06906"},{"key":"ref33","first-page":"1","article-title":"Proxylessnas: Direct neural architecture search on target task and hardware","author":"cai","year":"2019","journal-title":"Proc 7th Int Conf Learn Represent (ICLR)"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00169"},{"key":"ref31","article-title":"DARTS+: Improved differentiable architecture search with early stopping","author":"liang","year":"2019","journal-title":"arXiv 1909 06035"},{"key":"ref30","article-title":"FairNAS: Rethinking evaluation fairness of weight sharing neural architecture search","author":"chu","year":"2019","journal-title":"arXiv 1907 01845"},{"key":"ref37","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00257"},{"key":"ref36","first-page":"1","article-title":"Neural architecture search with Bayesian optimisation and optimal transport","author":"kandasamy","year":"2018","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref35","first-page":"459","article-title":"Neural optimizer search with reinforcement learning","author":"bello","year":"2017","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref34","first-page":"1","article-title":"SMASH: One-shot model architecture search through hypernetworks","author":"brock","year":"2018","journal-title":"Proc 6th Int Conf Learn Represent"},{"key":"ref60","first-page":"367","article-title":"Random search and reproducibility for neural architecture search","author":"li","year":"2020","journal-title":"Proc Uncertainty Artif Intell"},{"key":"ref62","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00207"},{"key":"ref61","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00139"},{"key":"ref63","first-page":"1","article-title":"Evaluating the search phase of neural architecture search","author":"yu","year":"2020","journal-title":"Proc 8th Int Conf Learn Represent (ICLR)"},{"key":"ref28","author":"shin","year":"2018","journal-title":"Differentiable neural network architecture search"},{"key":"ref64","first-page":"4074","author":"shirakawa","year":"2018","journal-title":"Dynamic Optimization of Neural Network Structures Using Probabilistic Modeling"},{"key":"ref27","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00492"},{"key":"ref65","article-title":"Towards fast adaptation of neural architectures with meta learning","author":"lian","year":"2020","journal-title":"Proc Int Conf Learn Represent"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.1109\/ICCVW.2019.00243"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00138"},{"key":"ref67","article-title":"ISTA-NAS: Efficient and consistent neural architecture search by sparse coding","author":"yang","year":"2020","journal-title":"arXiv 2010 06176"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01201"},{"key":"ref69","first-page":"493","article-title":"Asap: Architecture search, anneal and prune","author":"noy","year":"2020","journal-title":"Proc Int Conf Artif Intell Statist"},{"key":"ref2","first-page":"1","article-title":"Designing neural network architectures using reinforcement learning","author":"baker","year":"2017","journal-title":"Proc Int Conf Learn Represent (ICLR)"},{"key":"ref1","doi-asserted-by":"publisher","DOI":"10.1162\/106365602320169811"},{"key":"ref20","article-title":"DenseNet: Implementing efficient ConvNet descriptor pyramids","author":"iandola","year":"2014","journal-title":"Arxiv 1404 1869"},{"key":"ref22","article-title":"MobileNets: Efficient convolutional neural networks for mobile vision applications","author":"howard","year":"2017","journal-title":"arXiv 1704 04861"},{"key":"ref21","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00907"},{"key":"ref24","first-page":"2902","article-title":"Large-scale evolution of image classifiers","author":"real","year":"2017","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref23","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v33i01.33014780"},{"key":"ref26","first-page":"1","article-title":"Efficient architecture search by network transformation","volume":"32","author":"cai","year":"2018","journal-title":"Proc AAAI Conf Artif Intell"},{"key":"ref25","first-page":"4095","article-title":"Efficient neural architecture search via parameters sharing","author":"pham","year":"2018","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref50","first-page":"544","article-title":"Single path one-shot neural architecture search with uniform sampling","author":"guo","year":"2020","journal-title":"Proc Eur Conf Comput Vis"},{"key":"ref51","first-page":"1554","article-title":"Stabilizing differentiable architecture search via perturbation-based regularization","author":"chen","year":"2020","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref59","article-title":"Noisy differentiable architecture search","author":"chu","year":"2020","journal-title":"arXiv 2005 03566"},{"key":"ref58","article-title":"Adding gradient noise improves learning for very deep networks","author":"neelakantan","year":"2015","journal-title":"arXiv 1511 06807"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58555-6_28"},{"key":"ref56","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref55","article-title":"Mish: A self regularized non-monotonic neural activation function","author":"misra","year":"2019","journal-title":"arXiv 1908 08681"},{"key":"ref54","article-title":"Searching for activation functions","author":"ramachandran","year":"2018","journal-title":"Proc 6th Int Conf Learn Represent (ICLR)"},{"key":"ref53","first-page":"3","article-title":"Rectifier nonlinearities improve neural network acoustic models","volume":"30","author":"maas","year":"2013","journal-title":"Proc ICML"},{"key":"ref52","first-page":"807","author":"nair","year":"2010","journal-title":"Rectified Linear Units Improve Restricted Boltzmann Machines"},{"key":"ref10","doi-asserted-by":"publisher","DOI":"10.1016\/j.imavis.2019.06.005"},{"key":"ref11","article-title":"A comprehensive survey of neural architecture search: Challenges and solutions","volume":"abs 2006 2903","author":"ren","year":"2020","journal-title":"CoRR"},{"key":"ref40","first-page":"7827","article-title":"Neural architecture optimization","author":"luo","year":"2018","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref12","first-page":"1","article-title":"A study on encodings for neural architecture search","author":"white","year":"2020","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref13","article-title":"A survey on neural architecture search","author":"wistuba","year":"2019","journal-title":"arXiv 1905 01392"},{"key":"ref14","article-title":"Automated machine learning: State-of-the-art and open challenges","author":"shawi","year":"2019","journal-title":"arXiv 1906 02287"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.1016\/j.knosys.2020.106622"},{"key":"ref82","first-page":"876","article-title":"Data: Differentiable architecture approximation","volume":"32","author":"chang","year":"2019","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref16","article-title":"SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and ¡0.5MB model size","author":"iandola","year":"2016","journal-title":"arXiv 1602 07360"},{"key":"ref81","article-title":"Searching by generating: Flexible and efficient one-shot NAS with architecture generator","author":"huang","year":"2021","journal-title":"Proc IEEE Conf Comput Vis Pattern Recognit"},{"key":"ref17","article-title":"Places205-VGGNet models for scene recognition","author":"wang","year":"2015","journal-title":"arXiv 1508 01667"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/DICTA.2016.7797091"},{"key":"ref18","first-page":"1124","author":"ballester","year":"2016","journal-title":"On the performance of googlenet and alexnet applied to sketches"},{"key":"ref83","first-page":"759","article-title":"Network pruning via transformable architecture search","author":"dong","year":"2019","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref19","article-title":"ResNet in ResNet: Generalizing residual architectures","author":"targ","year":"2016","journal-title":"arXiv 1603 08029"},{"key":"ref80","doi-asserted-by":"publisher","DOI":"10.1007\/s00521-020-04915-6"},{"key":"ref4","author":"liu","year":"2019","journal-title":"DARTS Differentiable Architecture Search"},{"key":"ref3","author":"zoph","year":"2017","journal-title":"Neural architecture search with reinforcement learning"},{"key":"ref6","first-page":"1","article-title":"Understanding and robustifying differentiable architecture search","author":"zela","year":"2020","journal-title":"Proc Int Conf Learn Represent"},{"key":"ref5","author":"xu","year":"2020","journal-title":"Pc-Darts Partial Channel Connections for Memory-Efficient Architecture Search"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58548-8_46"},{"key":"ref8","first-page":"1","article-title":"SNAS: Stochastic neural architecture search","author":"xie","year":"2019","journal-title":"Proc 7th Int Conf Learn Represent (ICLR)"},{"key":"ref86","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2017.2773081"},{"key":"ref7","first-page":"7603","article-title":"Bayesnas: A Bayesian approach for neural architecture search","volume":"97","author":"zhou","year":"2019","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref49","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00378"},{"key":"ref9","first-page":"55:1","article-title":"Neural architecture search: A survey","volume":"20","author":"elsken","year":"2019","journal-title":"J Mach Learn Res"},{"key":"ref46","first-page":"6105","article-title":"EfficientNet: Rethinking model scaling for convolutional neural networks","author":"tan","year":"2019","journal-title":"Proc Int Conf Mach Learn"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00186"},{"key":"ref48","first-page":"1","article-title":"NAT: Neural architecture transformer for accurate and compact architectures","author":"guo","year":"2019","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref47","first-page":"1977","article-title":"XNAS: Neural architecture search with expert advice","author":"nayman","year":"2019","journal-title":"Proc Adv Neural Inf Process Syst (NeurIPS)"},{"key":"ref42","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00293"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01099"},{"key":"ref44","article-title":"Single-path NAS: Device-aware efficient ConvNet design","author":"stamoulis","year":"2019","journal-title":"arXiv 1905 04159"},{"key":"ref43","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01166"}],"container-title":["IEEE Access"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx7\/6287639\/9312710\/09461192.pdf?arnumber=9461192","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,12,17]],"date-time":"2021-12-17T19:56:41Z","timestamp":1639771001000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/9461192\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021]]},"references-count":86,"URL":"https:\/\/doi.org\/10.1109\/access.2021.3090918","relation":{},"ISSN":["2169-3536"],"issn-type":[{"value":"2169-3536","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021]]}}}