[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-450408-105":59,"doc-detail-450408-en":130},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":123,"head_meta":125,"extra_data":127,"updated_unix":129},105,"en","neural-architecture-search-using-network-embedding-and-generative-adversarial-networks","Neural architecture search using network embedding and generative adversarial networks","","Neural architecture search (NAS) is framed as an optimization problem for discovering high-performing network structures, but evaluating candidates is computationally expensive and supervised surrogate models require many labeled architectures. A surrogate-assisted swarm optimization method is proposed that uses network embedding to represent architectures in an unsupervised latent space. A generative adversarial network performs data augmentation to improve surrogate robustness under limited training data. Experiments on two NASBench search spaces show GNE-NAS achieves comparable or better performance than state-of-the-art NAS methods.",{"@graph":69,"@context":122},[70,84,105],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/neural-architecture-search-using-network-embedding-and-generative-adversarial-networks/450408/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/neural-architecture-search-using-network-embedding-and-generative-adversarial-networks/450408.png","ImageObject",300,407,{"name":92,"@type":93},"Connor ","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-06","2026-09-30",true,{"@type":102,"interactionType":103,"userInteractionCount":14},"InteractionCounter",{"@type":104},"ViewAction",{"@type":106,"mainEntity":107},"FAQPage",[108,114,118],{"name":109,"@type":110,"acceptedAnswer":111},"Why are surrogate models used in neural architecture search?","Question",{"text":112,"@type":113},"They forecast neural architecture performance so the method avoids expensive full training of each candidate during search.","Answer",{"name":115,"@type":110,"acceptedAnswer":116},"What problem does labeled data scarcity create for surrogate models?",{"text":117,"@type":113},"Supervised surrogate models still need large sets of well-trained, labeled architectures, which limits scalability and efficiency.",{"name":119,"@type":110,"acceptedAnswer":120},"How does GNE-NAS improve surrogate modeling when training data is limited?",{"text":121,"@type":113},"It embeds architectures using an unsupervised network embedding approach and augments surrogate-model training data via a generative adversarial network to enhance robustness and reduce reliance on many real evaluations.","https://schema.org",{"og:url":83,"og:type":124,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":126,"canonical":83},"index,follow",{"doc_id":128,"site_id":62},450408,1791323214,{"code":4,"msg":5,"data":131},{"doc_id":128,"user_id":132,"nickname":92,"user_avatar":133,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":134,"file_id":135,"file_url":136,"file_type":137,"file_size":138,"view_count":14,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":139,"language":140,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":141,"faqs":142,"seo_title":143,"seo_description":67,"update_tm":144,"read_time":145},687207022233,"https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d","[www. nature.com/scientificreports](www. nature.com/scientificreports)  \nOPEN  \nNeural architecture search using network embedding and generative adversarial networks  \nMortezaYousefi1, Vahid Mehrdad1􀀍 & Mohammad Bagher Dowlatshahi2  \nSurrogate models are used by recently proposed algorithms as a means of forecasting neural architecture performance. Rather than training the network from scratch, which speeds up evaluation of performance in the search for neural architecture. However, collecting a sufficient number of labeled architectures for training surrogate models is a time-consuming process. We suggest a surrogateassisted swarm optimization algorithm with network embedding for neural architecture search, as well as a generative adversarial networks for augmentation data (GNE-NAS) to improve the performance of surrogate models with limited training data. In this case, each architecture is meaningfully represented using an unsupervised learning technique. In the embedding space, architectures with a greater degree of structural similarity are positioned closer together. This proximity facilitates the training of surrogate models. Prior to training the surrogate model, we employ a generative adversarial network for data augmentation. This approach enhances the robustness of the surrogate model and concurrently reduces the need for a large number of real evaluations. The surrogate model achieves comparable or better performance when network embedding is applied, as demonstrated by experimental results on two distinct NASBench search spaces. Our proposed method, GNE-NAS, has been shown to outperform other state-of-the-art neural architecture search algorithms.  \nKeywords Neural architecture search (NAS), Particle swarm optimization (PSO), Generative adversarial networks (GAN), Surrogate model  \nDeep neural networks (DNNs) have achieved remarkable success in a wide range of artificial intelligence tasks, including image classification, object detection, and natural language processing1–3. Their performance depends not only on the optimization of network weights but also critically on the underlying architecture design. Traditionally, creating high-performing architectures has required substantial domain expertise and iterative experimentation. To reduce this dependence on manual design, Neural Architecture Search (NAS) has emerged as a powerful paradigm, formulating architecture design as an optimization problem over a predefined search space and aiming to discover network structures that maximize validation performance after training.  \nExisting NAS approaches can be broadly grouped into three categories: reinforcement learning RL-based, evolutionary algorithm EA-based, and gradient-based methods. RL-based NAS employs a controller that samples candidate architectures and learns to improve its proposals based on performance rewards. EA-based NAS treats architectures as individuals in a population, generating new ones via evolutionary operations like selection, crossover, and mutation. In contrast, gradient-based NAS relaxes the discrete nature of the search space into a continuous form, allowing architecture parameters and network weights to be optimized jointly via gradient descent. While EAs are known for their strong exploration capabilities4, their reliance on full training of each candidate makes them computationally costly, for instance, LargeEvo5 requires more than 3000 GPU-days for a single search.  \nSurrogate-Assisted Evolutionary Algorithms (SAEAs) address this computational bottleneck by replacing expensive full training with performance prediction through surrogate models. Examples include end-toend performance predictors6, adaptive switching between multiple surrogates7, and LSTM-based estimators8. Although such strategies reduce computational expense, supervised surrogate models still require large sets of well-trained architectures as labeled data, limiting their scalability. To address this, recent methods have ","cbCaioFQ9N6OE7yI","https://ap.wps.com/l/cbCaioFQ9N6OE7yI","pdf",4261539,22,"English","# Introduction\n# Background and Related Work\n## NAS methods: RL, EA, and gradient-based\n## Surrogate-assisted evolutionary algorithms and embedding approaches\n# Proposed Method\n## Graph2vec-based architecture embedding\n## GAN-based data augmentation for surrogate models\n# Experimental Evaluation","[{\"question\":\"Why are surrogate models used in neural architecture search?\",\"answer\":\"They forecast neural architecture performance so the method avoids expensive full training of each candidate during search.\"},{\"question\":\"What problem does labeled data scarcity create for surrogate models?\",\"answer\":\"Supervised surrogate models still need large sets of well-trained, labeled architectures, which limits scalability and efficiency.\"},{\"question\":\"How does GNE-NAS improve surrogate modeling when training data is limited?\",\"answer\":\"It embeds architectures using an unsupervised network embedding approach and augments surrogate-model training data via a generative adversarial network to enhance robustness and reduce reliance on many real evaluations.\"}]","Neural architecture search using network embedding and generative adversarial networks | PDF",1790733140,55]