[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122331-en":3,"doc-seo-122331-105":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},122331,2336464648746,"Skyler","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","HollerithEnergyML - A Prototype of a Machine Learning Energy Consumption Recommender System","Energy consumption impacts of machine learning classifiers directly affect sustainability research and practical deployment. Due to limited prior work in this area, a recommender system prototype is proposed to recommend energy consumption for different classifier options during training. The work presents a literature-driven problem formulation, design and development requirements, and an implemented prototype workflow. Future research directions are derived from the prototype demonstration and observed gaps in existing recommender approaches.","HollerithEnergyML  \nA Prototype of a Machine Learning Energy Consumption Recommender System  \nMichael Zanger1, Alexander Schulz1, Lukas Grodmeier1, Dion Agaj1, Rafael Schindler 1, Lukas Weiss1, and Michael Möhring1  \nAbstract: Energy consumption aspects of machine learning classifiers are important for research and practice as well. Due to sparse research in this area, a prototype of a recommender system was developed to provide energy consumption recommendations of different possible classifiers. The prototype is demonstrated as well as discussed and future research points are derived.  \nKeywords: AI, Energy Consumption, ML, Recommender  \n1 Introduction  \nThe usage of machine learning (ML) algorithms is important for organizations [Re20] . In general, ML is one part of artificial intelligence (AI) trying to identify patterns from data through the usage of different algorithms [RM19] . ML can generate business value and may improve the financial performance of enterprises [Re20] . Besides the benefits of using it, the usage of machine learning models can have a negative impact to climate change according to current research [Ka22] . The design, development and usage of machine learning models can consume a lot of energy [Ka22] . Global GHG emissions are associated with ML use and training [Ka22] . Therefore, it is important to know the impact of the usage of different machine learning algorithms to the energy consumption beforehand. However, there is sparse research regarding a literature review of the last years using methods recommended by e.g. Kitchenham [Ki09] . The literature review reveals a differentiated research landscape in the area of sustainable recommender systems for managing the resource consumption of ML-models. The conducted review reveals a gap: No existing recommender system specifically addresses the resource consumption of ML models. It is evident that a multitude offactors contribute to the energy usage [Ka22] of these models. However, the absence of a universally applicable formula to determine energy consumption adds a layer of complexity to the development of a resource-efficient recommender system. Therefore, this prototype demo-paper addresses this important gap and focusing on the research question: “How can a recommender system for analyzing the energy consumption of machine learning classifiers during training be implemented in a prototype? This research is part on  an ongoing research project. To answer the research  \n1 Reutlingen University, School of Informatics-HHZ, Alteburgstraße 150, Reutlingen, Germany, 72762 , michael.zanger@student.reutlingen-university.de  \ncba doi:10 . 18420/inf2024_ 132  \nquestion, we used a design science research approach according to Hevner [He04] as well as Peffers et al. [Pe07] and implemented the first steps of it. The addressed practical problem is the missing easy-to use energy recommendation system of the energy consumption of classifiers. The requirements of the solution as well as the design and development are described in section 2. Afterwards, the prototype is demonstrated (section 3) and finally a conclusion and future research steps for e.g., further evaluations are defined.  \n2 Development of the Recommender System Prototype  \nTo develop a prototype as an artifact [He04] [Pe07] and to provide an energy consumption recommendation for different classifiers during training phase (objective of the solution), different requirements had to be fulfilled. The following core requirements were collected based on different workshops running in 2023 : a) Usage of different classification algorithms and collection of related training energy consumption ; b) Recommendation based on user input of the number of numerical and categorial features as well as the sample size ; c) Web-based UI access to the recommender system. The following first classification algorithms were chosen to implement the first requirement (a) as recommended by the literature: Logistic Regressi","cbCaikRLJfehIoOD","https://ap.wps.com/l/cbCaikRLJfehIoOD","pdf",345005,1,5,"English","en",105,"# Introduction\n## Problem background and research gap\n# Development of the Recommender System Prototype\n## Requirements and chosen classifiers\n## Data collection and energy measurement\n## Prediction model and evaluation","[{\"question\":\"Why is energy consumption evaluation important for machine learning classifiers?\",\"answer\":\"Machine learning model design and training can consume substantial energy, contributing to greenhouse gas emissions. Knowing the impact of different classifier choices beforehand supports more resource-efficient development.\"},{\"question\":\"What does the proposed HollerithEnergyML prototype recommend?\",\"answer\":\"It recommends estimated energy consumption for different machine learning classifiers during the training phase. Recommendations are driven by user-provided counts of numerical and categorical features and the dataset size.\"},{\"question\":\"How is training energy consumption measured and predicted in the prototype?\",\"answer\":\"Energy during model execution is measured with the Python library CodeCarbon and recorded in kWh. A prediction model uses categorical feature count, numerical feature count, and row count to estimate energy consumption, with Random Forest Regressor showing the best performance in the described evaluation.\"}]","HollerithEnergyML - A Prototype of a Machine Learning Energy Consumption Recommender System | PDF",1785810051,13,{"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},"hollerithenergyml-a-prototype-of-a-machine-learning-energy-consumption-recommender-system","",{"@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/hollerithenergyml-a-prototype-of-a-machine-learning-energy-consumption-recommender-system/122331/",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","2026-08-04",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},"Why is energy consumption evaluation important for machine learning classifiers?","Question",{"text":75,"@type":76},"Machine learning model design and training can consume substantial energy, contributing to greenhouse gas emissions. Knowing the impact of different classifier choices beforehand supports more resource-efficient development.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What does the proposed HollerithEnergyML prototype recommend?",{"text":80,"@type":76},"It recommends estimated energy consumption for different machine learning classifiers during the training phase. Recommendations are driven by user-provided counts of numerical and categorical features and the dataset size.",{"name":82,"@type":73,"acceptedAnswer":83},"How is training energy consumption measured and predicted in the prototype?",{"text":84,"@type":76},"Energy during model execution is measured with the Python library CodeCarbon and recorded in kWh. A prediction model uses categorical feature count, numerical feature count, and row count to estimate energy consumption, with Random Forest Regressor showing the best performance in the described evaluation.","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},[93,97,101,105,109,114,119,122,127,130,134],{"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":21,"doc_module":4,"doc_module_name":46,"category_name":106,"show_sort_weight":107,"slug":108},"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":46,"category_name":111,"show_sort_weight":112,"slug":113},6,"Technology",50,"technology",{"id":115,"doc_module":4,"doc_module_name":46,"category_name":116,"show_sort_weight":117,"slug":118},7,"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":21,"slug":137},19,"General","general"]