[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117069-en":3,"doc-seo-117069-105":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},117069,3848291630094,"Emma Wilson","https://eur-avatar.wpscdn.com/davatar_085a072bc5b1113ac321206ff7593b45",8,"Research & Report","Aux-Drop - Handling Haphazard Inputs in Online Learning Using Auxiliary Dropouts","Many real-world online learning systems receive streaming data with haphazard structure: missing features, features becoming obsolete, sudden new features, and uncertainty about the total feature count. These conditions hinder training of learnable deep models, and existing deep-learning work rarely addresses the problem comprehensively. Aux-Drop introduces an auxiliary dropout regularization strategy adapted to the evolving feature space, minimizing disruption from chaotic feature changes, reducing co-adaptation, and lowering output dependence on auxiliary inputs across time-varying scenarios.","Aux-Drop: Handling Haphazard Inputs in Online Learning Using Auxiliary Dropouts  \nRohit Agarwal [agarwal.102497@gmail. com](agarwal.102497@gmail. com)  \n[Bio-AI Lab](Bio-AI Lab), [Department of Computer Science](Department of Computer Science)[ ](Department of Computer Science)UiT The Arctic University of Norway, Tromsø  \nDeepak Gupta  \nBio-AI Lab, Department of Computer Science UiT The Arctic University of Norway, Tromsø  \nAlexander Horsch  \nBio-AI Lab, Department of Computer Science UiT The Arctic University of Norway, Tromsø  \nDilip K. Prasad  \nBio-AI Lab, Department of Computer Science UiT The Arctic University of Norway, Tromsø  \nAbstract  \nMany real-world applications based on online learning produce streaming data that is haphazard in nature, i. e. , contains missing features, features becoming obsolete in time, the appearance of new features at later points in time and a lack of clarity on the total number of input features. These challenges make it hard to build a learnable system for such applications, and almost no work exists in deep learning that addresses this issue. In this paper, we present Aux-Drop, an auxiliary dropout regularization strategy for online learning that handles the haphazard input features in an effective manner. Aux-Drop adapts the conventional dropout regularization scheme for the haphazard input feature space ensuring that the final output is minimally impacted by the chaotic appearance of such features. It helps to prevent the co-adaptation of especially the auxiliary and base features, as well as reduces the strong dependence of the output on any of the auxiliary inputs of the model. This helps in better learning for scenarios where certain features disappear in time or when new features are to be modelled. The efficacy of Aux-Drop has been demonstrated through extensive numerical experiments on SOTA benchmarking datasets that include Italy Power Demand, HIGGS, SUSY and multiple UCI datasets. The code is available at [https://github.com/Rohit102497/Aux-Drop](https://github.com/Rohit102497/Aux-Drop).  \n1 Introduction  \nMany real-life applications produce streaming data that is difficult to model. Moreover, a lot of existing methods assume that the streaming data has a time-invariant fixed size and the models are trained accordingly (Gama, 2012; Nguyen et al., 2015) . However, this is not always true and the dimension of inputs can vary over time. The inputs can have missing data, missing features, obsolete features, sudden features and an unknown number of the total features. We define this here as the haphazard inputs. Formally, we define haphazard inputs as streaming data whose dimension varies at every time instance and there is no prior information about data received in the future. The characteristics of haphazard inputs are as follows: (1) Streaming data-Here data arrives sequentially and is modelled using online learning techniques. The model predicts an output based on the current data instance and then the actual output is revealed. The model gets trained based on the loss from its prediction and the actual output, and this updated model is used  \nTable 1: Comparison of different online deep learning models with respect to the characteristics of the haphazard inputs (C1-C6) . We showcase the inability of online deep learning methods in handling haphazard inputs even when other techniques like imputation, extrapolation, priori information and Gaussian noise are employed.  \n\n| Characteristics | Aux-Drop | Online Deep\u003Cbr>Learning\u003Cbr>Methods\u003Cbr>like\u003Cbr>ODL | ODL\u003Cbr>+\u003Cbr>Online\u003Cbr>Data\u003Cbr>Imputation | ODL\u003Cbr>+\u003Cbr>Extrapolation | ODL\u003Cbr>+\u003Cbr>Prior\u003Cbr>Information | ODL\u003Cbr>+\u003Cbr>Gaussian\u003Cbr>Noise |\n| --- | --- | --- | --- | --- | --- | --- |\n| Streaming data (C1) | ✓ | ✓ | ✓ | ✓ | ✓ | ✓ |\n| Missing data (C2) | ✓ | × | ✓ | × | ✓ | ✓ |\n| Missing features (C3) | ✓ | × | × | × | × | ✓ |\n| Obsolete features (C4) | ✓ | × | × | ✓ | × | ✓ |\n| Sudden features (C5) | ✓ | × | × | × | × | × |\n| Unknown ","cbCaikbYtMqlF7OI","https://ap.wps.com/l/cbCaikbYtMqlF7OI","pdf",942782,1,21,"English","en",105,"# Abstract\n# 1 Introduction\n## Problem definition: haphazard inputs\n## Challenges and limitations of existing methods\n## Comparison with baseline online deep learning approaches","[{\"question\":\"What are haphazard inputs in online learning?\",\"answer\":\"Haphazard inputs are streaming data whose input dimension changes at every time instance, with no prior knowledge about future arrivals. They include missing data, missing or obsolete features, sudden new features, and an unknown total number of features.\"},{\"question\":\"How does Aux-Drop handle haphazard feature spaces?\",\"answer\":\"Aux-Drop adapts conventional dropout to the auxiliary and evolving feature setting, ensuring the final output is minimally impacted by chaotic appearance and disappearance of features. It also reduces co-adaptation between auxiliary and base features and lowers dependence on auxiliary inputs.\"},{\"question\":\"How is Aux-Drop evaluated?\",\"answer\":\"Aux-Drop is tested through extensive numerical experiments on SOTA benchmarking datasets, including Italy Power Demand, HIGGS, SUSY, and multiple UCI datasets. Results support its effectiveness under time-varying feature conditions.\"}]","Aux-Drop - Handling Haphazard Inputs in Online Learning Using Auxiliary Dropouts | PDF",1785673540,53,{"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},"aux-drop-handling-haphazard-inputs-in-online-learning-using-auxiliary-dropouts","",{"@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/aux-drop-handling-haphazard-inputs-in-online-learning-using-auxiliary-dropouts/117069/",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 are haphazard inputs in online learning?","Question",{"text":75,"@type":76},"Haphazard inputs are streaming data whose input dimension changes at every time instance, with no prior knowledge about future arrivals. They include missing data, missing or obsolete features, sudden new features, and an unknown total number of features.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does Aux-Drop handle haphazard feature spaces?",{"text":80,"@type":76},"Aux-Drop adapts conventional dropout to the auxiliary and evolving feature setting, ensuring the final output is minimally impacted by chaotic appearance and disappearance of features. It also reduces co-adaptation between auxiliary and base features and lowers dependence on auxiliary inputs.",{"name":82,"@type":73,"acceptedAnswer":83},"How is Aux-Drop evaluated?",{"text":84,"@type":76},"Aux-Drop is tested through extensive numerical experiments on SOTA benchmarking datasets, including Italy Power Demand, HIGGS, SUSY, and multiple UCI datasets. Results support its effectiveness under time-varying feature conditions.","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"]