[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117465-en":3,"doc-seo-117465-105":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},117465,1374391975076,"Riley","https://ap-avatar.wpscdn.com/avatar/14000253ca4ec9f6853?x-image-process=image/resize,m_fixed,w_180,h_180&k=1783305029341752051",8,"Research & Report","Zero-Waste Machine Learning - A Research Path for Efficient, Resource-Recycling Models","Machine learning systems, especially artificial neural networks, increasingly require high computation and energy during training, driving long processing times and substantial carbon footprints across large-scale deployments. This work reframes efficiency through “zero-waste machine learning,” emphasizing reuse of already available computations, runtime-accessible partial information, and knowledge gained from prior training in continually learned models. The paper investigates related questions on learning with less data, selecting relevant samples efficiently, and building on previously trained models to reduce further training needs, and summarizes research directions and early results for this approach.","ECAI 2024  \nU. Endriss et al. (Eds.)  \n© 2024 The Authors.  \nThis article is published online with Open Access by IOS Press and distributed under the terms of the Creative Commons Attribution Non-Commercial License 4.0 (CC BY-NC 4.0). doi:10.3233/FAIA240466  \n43  \nZero-Waste Machine Learning  \nTomasz Trzcinskia,b, * , Bartłomiej Twardowski a,d,e, Bartosz Zieliskia,c, Kamil Adamczewskia and  \nBartosz Wójcika,c  \naIDEAS NCBR  \nb Warsaw University of Technology  \ncJagiellonian University  \nd Computer Vision Center, Barcelona, Spain  \neUniversitat Autonoma de Barcelona, Barcelona, Spain  \nORCID (Tomasz Trzcinski): [https://orcid.org/0000-0002-1486-8906](https://orcid.org/0000-0002-1486-8906), ORCID (Bartłomiej Twardowski): [https://orcid.org/0000-0003-2117-8679](https://orcid.org/0000-0003-2117-8679), ORCID (Bartosz Zieli´nski): https://orcid.org/0000-0002-3063-3621, ORCID (Kamil Adamczewski): [https://orcid.org/0000-0002-2917-4392](https://orcid.org/0000-0002-2917-4392), ORCID (Bartosz Wójcik):  \n[https://orcid.org/0000-0002-1100-4176](https://orcid.org/0000-0002-1100-4176)  \nAbstract. Today, both science and industry rely heavily on machine learning models, predominantly artificial neural networks, that become increasingly complex and demand more computing resources to be trained. In this paper, we will look holistically at the efficiency of machine learning models and draw the inspirations to address their main challenges from the green sustainable economy principles. Instead of constraining some computations or memory used by the models, we will focus on reusing what is available to them: computations done in the previous processing steps, partial information accessible at run-time, or knowledge gained by the model during previous training sessions in continually learned models. This new research path of zero-waste machine learning can lead to several research questions related to efficiency of contemporary neural networks-how machine learning models can learn better with less data? How they select relevant data samples out of many? Finally, how can they build on top of already trained models to reduce the need for more training samples? Here, we explore all the above questions and attempt to answer them.  \n1 Introduction  \nToday, both science and industry heavily depend on machine learning models, especially artificial neural networks, which are becoming increasingly complex and require substantial computational resources. This trend is evident in a wide range of applications, from medical image processing to robotics. However, the most significant computational demands of machine learning models are seen in large high-energy physics experiments, where an enormous amount of data is generated and analyzed. For example, the ALICE experiment at CERN’s Large Hadron Collider (LHC), the world’s largest and most powerful particle accelerator, gathers several petabytes of data every hour, which is processed by numerous machine learning models.  \nThe computations run by machine learning models to process this increasing amount of data come at an enormous price of long processing time, high energy consumption and large carbon footprint generated by the computational infrastructure [32]. Existing ap-  \n∗ Corresponding Author. Email: [tomasz.trzcinski@ideas-ncbr.pl](tomasz.trzcinski@ideas-ncbr.pl)  \nproaches to reduce this burden are either focused on constraining the optimization with a limited budget of computational resources [17] or they attempt to compress models [15] .  \nIn this paper, we look holistically at the efficiency of machine learning models and draw inspiration to address their main challenges from the green sustainable economy principles. Instead of limiting training of machine learning models, we ask a different question: how can we make the best out of the resources and information and computations that we already have access to? Instead of constraining the number of computations or memory used by the models, we focus ","cbCaiuPYvYGKp6bo","https://ap.wps.com/l/cbCaiuPYvYGKp6bo","pdf",1241469,1,7,"English","en",105,"# Introduction\n## Efficiency challenges in large-scale ML\n## Zero-waste machine learning research questions\n## Definition and focus areas\n## Real-life use cases and outlook","[{\"question\":\"What is “zero-waste machine learning” in this paper?\",\"answer\":\"It is a research direction aimed at more efficient machine learning by reusing available resources—especially conditioning computations and acquiring knowledge in continually trained models.\"},{\"question\":\"Why does the paper emphasize efficiency and sustainability?\",\"answer\":\"Training and inference at scale consume significant compute resources, leading to long processing times, high energy use, and large carbon footprints in supporting infrastructure.\"},{\"question\":\"Which research questions does the paper address?\",\"answer\":\"It asks how to optimize resource usage via a zero-waste policy and computation recycling, how to train efficient neural representations through conditioned computations, how to accumulate knowledge in and beyond continual learning, and how to evaluate efficiency from the recycling perspective.\"}]","Zero-Waste Machine Learning - A Research Path for Efficient, Resource-Recycling Models | PDF",1785675986,18,{"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},"zero-waste-machine-learning-a-research-path-for-efficient-resource-recycling-models","",{"@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/zero-waste-machine-learning-a-research-path-for-efficient-resource-recycling-models/117465/",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 is “zero-waste machine learning” in this paper?","Question",{"text":75,"@type":76},"It is a research direction aimed at more efficient machine learning by reusing available resources—especially conditioning computations and acquiring knowledge in continually trained models.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why does the paper emphasize efficiency and sustainability?",{"text":80,"@type":76},"Training and inference at scale consume significant compute resources, leading to long processing times, high energy use, and large carbon footprints in supporting infrastructure.",{"name":82,"@type":73,"acceptedAnswer":83},"Which research questions does the paper address?",{"text":84,"@type":76},"It asks how to optimize resource usage via a zero-waste policy and computation recycling, how to train efficient neural representations through conditioned computations, how to accumulate knowledge in and beyond continual learning, and how to evaluate efficiency from the recycling perspective.","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},[93,97,101,105,110,115,119,122,127,130,134],{"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":21,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":120,"slug":121},30,"research-report",{"id":123,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},9,"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]