[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-121631-en":3,"doc-seo-121631-105":30,"detail-sidebar-cat-0-en-105":92},{"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":20,"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},121631,1099513958607,"Jiven","https://ap-avatar.wpscdn.com/avatar/100002390cf8733938c?x-image-process=image/resize,m_fixed,w_180,h_180&k=1778829742770036399",8,"Research & Report","Interpretable machine learning models for predicting with missing values - Thesis","Machine learning models are frequently applied when input values are missing at either training time or prediction time. Improper handling of missingness can introduce bias or produce models that are unusable in practice without imputation, while imputation for complex functions is often difficult to interpret. This thesis studies test-time incomplete data and uses interpretable models so humans can understand predictions. It first addresses recurrent missing-variable patterns with small per-pattern sample sizes by proposing SPSM with shared coefficients to avoid reliance on imputation and to yield sparse descriptive interpretations; it then studies non-pattern missingness via MINTY, a sparse linear rule model that balances fit and interpretability while learning replacement variables.","Thesis for The Degree of Licentiate of Engineering  \nInterpretable machine learning models for predicting with missing values  \nLena Stempfle  \nDepartment of Computer Science and Engineering Chalmers University of Technology | University of Gothenburg  \nGothenburg, Sweden, 2023  \nInterpretable machine learning models for predicting with missing values  \nLena Stempfle  \n© Lena Stempfle, 2023  \nexcept where otherwise stated.  \nAll rights reserved.  \nISSN 1652-876X  \nDepartment of Computer Science and Engineering Division of Data Science and AI  \nChalmers University of Technology | University of Gothenburg SE-412 96 G¨oteborg,  \nSweden  \nPhone: +46(0)31 772 1000  \nPrinted by Chalmers Digitaltryck, Gothenburg, Sweden 2023 .  \nInterpretable machine learning models for predicting with missing values  \nLena Stempfle  \nDepartment of Computer Science and Engineering  \nChalmers University of Technology | University of Gothenburg  \nAbstract  \nMachine learning models are often used in situations where model inputs are missing either during training or at the time of prediction. If missing values are not handled appropriately, they can lead to increased bias or to models that are not applicable in practice without imputing the values of the unobserved variables. However, the imputation of missing values is often inadequate and difficult to interpret for complex imputation functions.  \nIn this thesis, we focus on predictions in the presence of incomplete data at test time, using interpretable models that allow humans to understand the predictions. Interpretability is especially necessary when important decisions are at stake, such as in healthcare. First, we investigate, the situation where variables are missing in recurrent patterns and sample sizes are small per pattern. We propose SPSM that allows coefficient sharing between a main model and pattern submodels in order to make efficient use of data and tobe independent on imputation. To enable interpretability, the model can be expressed as a short description introduced by sparsity. Then, we explore situations where missingness does not occur in patterns and suggest the sparse linear rule model MINTY that naturally trades off between interpretability and the goodness of fit while being sensitive to missing values at test time. To this end, we learn replacement variables, indicating which features in a rule can be alternatively used when the original feature was not measured, assuming some redundancy in the covariates.  \nOur results have shown that the proposed interpretable models can be used for prediction with missing values, without depending on imputation. We conclude that more work can be done in evaluating interpretable machine learning models in the context of missing values at test time.  \nKeywords  \nMachine learning, interpretable machine learning, missing values, healthcare  \nList of Publications  \nThis thesis is based on the following appended publications:  \n[Paper I] H´akon Valur Dansson, Lena Stempfle, Hildur Egilsd´ottir, Alexander Schliep, Erik Portelius, Kaj Blennow, Henrik Zetterberg and Fredrik D. Johansson for the Alzheimer’s Disease Neuroimaging Initiative (ADNI), Predicting progression and cognitive decline in amyloid-positive patients with Alzheimer’s disease  \nAlzheimer’s Research and Therapy 13, 151 (2021) .  \n[Paper II] Lena Stempfle, Ashkan Panahi, Fredrik D. Johansson, Sharing pattern submodels for prediction with missing values  \nThirty-Seventh Conference on Artificial Intelligence (AAAI-23, To appear)(2023) .  \nThe following manuscript has been produced during my Ph.D. studies but has not been published yet.  \n[Paper III] Lena Stempfle, Fredrik D. Johansson, Learning replacement variables in interpretable rule-based models  \nManuscript in preparation.  \nAcknowledgment  \nFirst of all, I would like to thank my supervisor Fredrik Johansson for his never-ending support and inspiring leadership. This thesis and our results would not exist without your guidance, k","cbCaifC1sUQZPOm9","https://ap.wps.com/l/cbCaifC1sUQZPOm9","pdf",1607028,1,78,"English","en",105,"# Abstract\n# Keywords\n# List of Publications\n# Acknowledgment\n# Contents","[{\"question\":\"Why is missing-value handling important for machine learning predictions?\",\"answer\":\"Missing values can increase bias or make models impractical unless missingness is handled correctly. The thesis emphasizes that imputation may also be inadequate and hard to interpret for complex functions.\"},{\"question\":\"How does the thesis ensure interpretability when predicting with incomplete data?\",\"answer\":\"It uses interpretable models designed to let humans understand predictions. Interpretability is achieved through sparse descriptive representations and rule-based structures.\"},{\"question\":\"What approaches are proposed for different missingness scenarios?\",\"answer\":\"For recurrent missing-variable patterns with small sample sizes, the thesis proposes SPSM with shared coefficients between a main model and pattern submodels. For missingness that does not occur in patterns, it proposes MINTY, a sparse linear rule model that learns replacement variables to handle unmeasured features at test time.\"}]","Interpretable machine learning models for predicting with missing values - Thesis | PDF",1785805837,197,{"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":87,"head_meta":89,"extra_data":91,"updated_unix":28},"interpretable-machine-learning-models-for-predicting-with-missing-values-thesis","",{"@graph":36,"@context":86},[37,54,69],{"@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/interpretable-machine-learning-models-for-predicting-with-missing-values-thesis/121631/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":63,"encodingFormat":61,"isAccessibleForFree":64,"interactionStatistic":65},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-05","2026-08-04",true,{"@type":66,"interactionType":67,"userInteractionCount":20},"InteractionCounter",{"@type":68},"ViewAction",{"@type":70,"mainEntity":71},"FAQPage",[72,78,82],{"name":73,"@type":74,"acceptedAnswer":75},"Why is missing-value handling important for machine learning predictions?","Question",{"text":76,"@type":77},"Missing values can increase bias or make models impractical unless missingness is handled correctly. The thesis emphasizes that imputation may also be inadequate and hard to interpret for complex functions.","Answer",{"name":79,"@type":74,"acceptedAnswer":80},"How does the thesis ensure interpretability when predicting with incomplete data?",{"text":81,"@type":77},"It uses interpretable models designed to let humans understand predictions. Interpretability is achieved through sparse descriptive representations and rule-based structures.",{"name":83,"@type":74,"acceptedAnswer":84},"What approaches are proposed for different missingness scenarios?",{"text":85,"@type":77},"For recurrent missing-variable patterns with small sample sizes, the thesis proposes SPSM with shared coefficients between a main model and pattern submodels. For missingness that does not occur in patterns, it proposes MINTY, a sparse linear rule model that learns replacement variables to handle unmeasured features at test time.","https://schema.org",{"og:url":52,"og:type":88,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":90,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":93},[94,98,102,106,111,116,121,124,129,132,136],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":95,"show_sort_weight":96,"slug":97},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":99,"show_sort_weight":100,"slug":101},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":103,"show_sort_weight":104,"slug":105},"Exam",70,"exam",{"id":107,"doc_module":4,"doc_module_name":46,"category_name":108,"show_sort_weight":109,"slug":110},5,"Comic",60,"comic",{"id":112,"doc_module":4,"doc_module_name":46,"category_name":113,"show_sort_weight":114,"slug":115},6,"Technology",50,"technology",{"id":117,"doc_module":4,"doc_module_name":46,"category_name":118,"show_sort_weight":119,"slug":120},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":122,"slug":123},30,"research-report",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":126,"show_sort_weight":127,"slug":128},9,"Religion & Spirituality",20,"religion-spirituality",{"id":127,"doc_module":4,"doc_module_name":46,"category_name":130,"show_sort_weight":127,"slug":131},"World Cup","world-cup",{"id":133,"doc_module":4,"doc_module_name":46,"category_name":134,"show_sort_weight":133,"slug":135},10,"Lifestyle","lifestyle",{"id":137,"doc_module":4,"doc_module_name":46,"category_name":138,"show_sort_weight":107,"slug":139},19,"General","general"]