[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117484-en":3,"doc-seo-117484-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},117484,1099514067415,"Rowan","https://ap-avatar.wpscdn.com/avatar/100002539d78ffe74a7?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779092875211072502",8,"Research & Report","A Scalable and Efficient Iterative Method for Copying Machine Learning Classifiers","Differential replication through copying transfers a trained machine-learning model’s decision behavior to a new model with stronger or better-suited features under changing industrial constraints. The paper addresses the limitations of prior single-pass copying by proposing a sequential, iterative method that reduces the computational resources needed to train or maintain a copy. Experiments on synthetic and real-world datasets demonstrate reduced time and resource consumption while maintaining or improving predictive accuracy, lowering ongoing maintenance costs for production systems.","A Scalable and E􀀎cient Iterative Method for Copying Machine Learning Classi􀀌ers  \nNahuel Statuto [nahuel.statuto@esade.edu](nahuel.statuto@esade.edu)  \nDepartment of Operations, Innovation and Data Sciences Universitat Ramon Llull, ESADE  \nSant Cugat del Vall􀀒es, 08172, Catalonia, Spain  \nIrene Unceta [irene.unceta@esade.edu](irene.unceta@esade.edu)  \nDepartment of Operations, Innovation and Data Sciences Universitat Ramon Llull, ESADE  \nSant Cugat del Vall􀀒es, 08172, Catalonia, Spain  \nJordi Nin [jordi.nin@esade.edu](jordi.nin@esade.edu)  \nDepartment of Operations, Innovation and Data Sciences Universitat Ramon Llull, ESADE  \nSant Cugat del Vall􀀒es, 08172, Catalonia, Spain  \nOriol Pujol oriol [pujol@ub.edu](pujol@ub.edu)  \nDepartament de Matem􀀒atiques i Inform􀀒atica Universitat de Barcelona  \nBarcelona, 08007, Catalonia, Spain  \nEditor: Maya Gupta  \nAbstract  \nDi􀀋erential replication through copying refers to the process of replicating the decision behavior of a machine learning model using another model that possesses enhanced features and attributes. This process is relevant when external constraints limit the performance of an industrial predictive system. Under such circumstances, copying enables the retention of original prediction capabilities while adapting to new demands. Previous research has focused on the single-pass implementation for copying. This paper introduces a novel sequential approach that signi􀀌cantly reduces the amount of computational resources needed to train or maintain a copy, leading to reduced maintenance costs for companies using machine learning models in production. The e􀀋ectiveness of the sequential approach is demonstrated through experiments with synthetic and real-world datasets, showing significant reductions in time and resources, while maintaining or improving accuracy.  \nKeywords: Sustainable AI, transfer learning, environmental adaptation, optimization, and model enhancement.  \n1. Introduction  \nMachine learning has become widespread in many industries. Supervised algorithms have been shown to automate tasks with a higher precision and lower costs when compared to previous methods (G􀀓omez et al., 2018; Gharibshah and Zhu, 2021; Shehab et al., 2022; Feizabadi, 2022) .  \n􀀍c2023 Nahuel Statuto, Irene Unceta, Jordi Nin, and Oriol Pujol.  \nLicense: CC-BY 4.0, see [https://creativecommons.org/licenses/by/4.0/](https://creativecommons.org/licenses/by/4.0/. Attribution)[. Attribution](https://creativecommons.org/licenses/by/4.0/. Attribution) requirements are provided  \nat [http://jmlr.org/papers/v24/23-0135.html](http://jmlr.org/papers/v24/23-0135.html).  \nStatuto, Unceta, Nin, and Pujol  \nHowever, maintaining the performance of industrial machine learning models requires constant monitoring. The environment where these models are deployed is subject to continuous change due to factors that are both internal and external to the companies implementing them. These factors may include new production needs, technological updates, novel market trends, or regulatory amendments. Neglecting these changes can result in model degradation and decreased performance. In the worst-case scenario, failure to adapt to these changes can render a model obsolete. To prevent this, regular model monitoring is essential in any industrial machine learning application. Once deployed in production, models are frequently checked for signs of performance deviation, which can occur as early as a few months after deployment. When such signs appear, models are either fully or partially retrained and substituted (Wu et al., 2020) . Unfortunately, this process is both time-consuming and costly. The nature of modern model architectures is complex. Hence, the task of managing and updating an existing model demands signi􀀌cant computational resources (Chen, 2019) . Handling multiple models simultaneously poses a signi􀀌cant challenge for companies. Thus, ensuring long-term sustainability remains one of the main obstacles faced by ","cbCailUd0FmGbV9n","https://ap.wps.com/l/cbCailUd0FmGbV9n","pdf",716354,1,34,"English","en",105,"# Introduction\n## Motivation: model monitoring and adaptation\n## Copying via differential replication\n## Problem framing and limitations of single-pass methods\n# Proposed iterative approach","[{\"question\":\"What problem does differential replication through copying address?\",\"answer\":\"It replicates a model’s decision behavior using another model with enhanced characteristics when external constraints limit the original system’s performance.\"},{\"question\":\"How does the paper improve over previous single-pass copying methods?\",\"answer\":\"It introduces a novel sequential/iterative approach that reduces computational resources required to train or maintain a copied model.\"},{\"question\":\"What evidence supports the effectiveness of the proposed method?\",\"answer\":\"Experiments on synthetic and real-world datasets show significant reductions in time and resources while maintaining or improving accuracy.\"}]","A Scalable and Efficient Iterative Method for Copying Machine Learning Classifiers | PDF",1785676121,86,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"a-scalable-and-efficient-iterative-method-for-copying-machine-learning-classifiers","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/a-scalable-and-efficient-iterative-method-for-copying-machine-learning-classifiers/117484/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-02",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does differential replication through copying address?","Question",{"text":75,"@type":76},"It replicates a model’s decision behavior using another model with enhanced characteristics when external constraints limit the original system’s performance.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the paper improve over previous single-pass copying methods?",{"text":80,"@type":76},"It introduces a novel sequential/iterative approach that reduces computational resources required to train or maintain a copied model.",{"name":82,"@type":73,"acceptedAnswer":83},"What evidence supports the effectiveness of the proposed method?",{"text":84,"@type":76},"Experiments on synthetic and real-world datasets show significant reductions in time and resources while maintaining or improving accuracy.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]