[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119036-en":3,"doc-seo-119036-105":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},119036,137441390410,"Hazel","https://ap-avatar.wpscdn.com/avatar/2000252f4ab5702993?_k=1776741390130283984",8,"Research & Report","Automated Machine Learning for Positive-Unlabelled Learning - Overview","Positive-Unlabelled (PU) learning tackles the task of building classifiers from data containing labelled positive instances and unlabelled instances whose true label is unknown. The work extends an earlier Auto-ML approach, GA-Auto-PU, by proposing two new PU-focused systems: BO-Auto-PU using Bayesian optimisation and EBO-Auto-PU combining evolutionary search with Bayesian optimisation. An extensive experimental evaluation compares all three Auto-ML systems against each other and established PU methods across 60 datasets.","arXiv :2401 .06452v1 [ cs .LG] 12 Jan 2024  \nAutomated Machine Learning for Positive-Unlabelled Learning  \nAutomated Machine Learning for Positive-Unlabelled Learning  \nJack D. Saunders [jsaunders0027@gmail.com](jsaunders0027@gmail.com)  \nSchool of Computing, University of Kent Canterbury, CT2 7NT, UK  \nAlex A. Freitas [A.A.Freitas@kent.ac.uk](A.A.Freitas@kent.ac.uk)  \nSchool of Computing, University of Kent Canterbury, CT2 7NT, UK  \nEditor: Editor Name  \nAbstract  \nPositive-Unlabelled (PU) learning is a growing field of machine learning that aims to learn classifiers from data consisting of labelled positive and unlabelled instances, which can be in reality positive or negative, but whose label is unknown. An extensive number of methods have been proposed to address PU learning over the last two decades, so many so that selecting an optimal method for a given PU learning task presents a challenge. Our previous work has addressed this by proposing GA-Auto-PU, the first Automated Machine Learning (Auto-ML) system for PU learning. In this work, we propose two new AutoML systems for PU learning: BO-Auto-PU, based on a Bayesian Optimisation approach, and EBO-Auto-PU, based on a novel evolutionary/Bayesian optimisation approach. We also present an extensive evaluation of the three Auto-ML systems, comparing them to each other and to well-established PU learning methods across 60 datasets (20 real-world datasets, each with 3 versions in terms of PU learning characteristics) .  \nKeywords: Positive-Unlabelled Learning, Automated Machine Learning (Auto-ML), Bayesian Optimisation, Genetic Algorithm, Classification  \n1. Introduction  \nPositive-Unlabelled (PU) learning is a field of machine learning that aims to learn classifiers from data consisting of a set of labelled positive instances and a set of unlabelled instances which can be either positive or negative, but whose label is unknown (Bekker and Davis, 2020) . To accurately utilise a standard machine learning method, a fully labelled dataset is required. However, in real-world applications this is often unfeasible due to the expense or impracticality of obtaining fully labelled data (Bekker and Davis, 2020) . To address this issue, many methods have been proposed to learn classifiers from partially labelled data (Bekker and Davis, 2020) (Jaskie and Spanias, 2019) . This field is referred to as semisupervised classification and usually focuses on learning classifiers from data where only a subset of instances is labelled as the positive or negative class, whilst all the other instances (usually the majority of instances) are unlabelled. PU learning is a specialised case of semisupervised learning where, among the truly positive-class instances, only a subset is labelled as positive; whilst all the other instances are unlabelled – i.e., it is unclear if they are positive or negative instances (Bekker and Davis, 2020) . Thus, positive-unlabelled learning presents  \nSaunders and Freitas  \nan arguably greater challenge than the standard semi-supervised learning paradigm, due to the complete absence of negative labels in the data.  \nThere are many real world applications of PU learning, including cybersecurity (Zhang et al., 2017) (Luo et al., 2018), bioinformatics (Yang et al. , 2012) (Nikdelfaz and Jalili, 2018) (Vasighizaker and Jalili, 2018), and text mining (Liu et al., 2002) (Ke et al., 2012)(Liu and Peng, 2014) . For example, PU learning has previously been used for predicting disease-related genes (Nikdelfaz and Jalili, 2018) . This is a PU learning task where diseaseassociated genes are the labelled positive instances, as confirmed by biomedical experiments. However, the vast majority of the genes thought not to be associated with diseases have not undergone such experiments, since these experiments are expensive. As such, the genes without association with diseases are better thought of as unlabelled instances as there is no experimental evidence indicating either association o","cbCainzUZJIjVtqe","https://ap.wps.com/l/cbCainzUZJIjVtqe","pdf",521802,1,36,"English","en",105,"# Introduction\n## PU learning problem setting\n## Motivation and real-world applications\n## Two-step PU learning approach\n## Prior Auto-ML work (GA-Auto-PU) and limitations","[{\"question\":\"What is Positive-Unlabelled (PU) learning?\",\"answer\":\"PU learning learns classifiers from labelled positive instances and unlabelled instances that may be positive or negative, with the unknown label treated as unlabeled.\"},{\"question\":\"Why is PU learning more challenging than standard semi-supervised learning?\",\"answer\":\"Because the data lacks negative labels entirely, making it harder to separate positive from negative cases using incomplete supervision.\"},{\"question\":\"What Auto-ML systems are proposed for PU learning in this work?\",\"answer\":\"BO-Auto-PU uses Bayesian optimisation, while EBO-Auto-PU combines an evolutionary component with Bayesian optimisation; both are compared alongside the earlier GA-Auto-PU system.\"}]","Automated Machine Learning for Positive-Unlabelled Learning - Overview | PDF",1785722033,91,{"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},"automated-machine-learning-for-positive-unlabelled-learning-overview","",{"@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/automated-machine-learning-for-positive-unlabelled-learning-overview/119036/",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","2026-08-03",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 Positive-Unlabelled (PU) learning?","Question",{"text":75,"@type":76},"PU learning learns classifiers from labelled positive instances and unlabelled instances that may be positive or negative, with the unknown label treated as unlabeled.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Why is PU learning more challenging than standard semi-supervised learning?",{"text":80,"@type":76},"Because the data lacks negative labels entirely, making it harder to separate positive from negative cases using incomplete supervision.",{"name":82,"@type":73,"acceptedAnswer":83},"What Auto-ML systems are proposed for PU learning in this work?",{"text":84,"@type":76},"BO-Auto-PU uses Bayesian optimisation, while EBO-Auto-PU combines an evolutionary component with Bayesian optimisation; both are compared alongside the earlier GA-Auto-PU system.","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"]