{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T15:46:23Z","timestamp":1791906383177,"version":"4.3.4"},"reference-count":48,"publisher":"JMIR Publications Inc.","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JMIR AI"],"abstract":"Abstract<\/jats:title>\n \n Background<\/jats:title>\n Measles is a major cause of death among infants despite free routine immunization (RI) in low- and middle-income countries (LMICs). Vaccination defaulting contributes to measles outbreak occurrences. Conventional immunization defaulter identification methods are expensive to implement, prone to data falsification, reactive, and overestimate uptake. Several machine learning studies leveraged the use of child and maternal characteristics as predictors of immunization defaulting, overlooking community, demographic, and socioeconomic factors. Similarly, there is limited research regarding the use of interpretable deep learning and advanced ensemble methods to improve model performance in predicting RI defaulters and establishing significant predictors at the national level in the LMIC context, like Uganda.<\/jats:p>\n <\/jats:sec>\n \n Objective<\/jats:title>\n This study predicts infants at a risk of defaulting on measles vaccination in Uganda and Kenya using advanced ensemble learning architectures.<\/jats:p>\n <\/jats:sec>\n \n Methods<\/jats:title>\n \n The study benchmarked classical and deep learning models using Demographic and Health Survey (DHS) datasets (2016-2022) with child, maternal, community, social-economic, and demographic characteristics as predictors. The robust modeling pipeline began with feature selection according to the literature and the expert domain. It was followed by vaccination eligibility, missing value handling, and feature elimination using the phi-k correlation framework. To mitigate class imbalance, synthetic minority over-sampling technique for nominal continuous features was used in a training pipeline after the RobustScaler. Nine baseline classifiers, including Tabular Network (TabNet), multilayer perceptron (MLP), random forest (RF), gradient boosting (GB), adaptive-boosting (AB), among others, underwent stratified k-fold cross-validation and hyperparameter optimization via grid search. Custom soft voting was evaluated alongside TabNet stack architecture for superior ensemble prediction. Models were evaluated for accuracy, precision, recall, and\n F<\/jats:italic>\n 1<\/jats:sub>\n -score. Model explainability was decoded using Shapley additive explanations (SHAP), while Streamlit and SQLite formed the tool architecture.\n <\/jats:p>\n <\/jats:sec>\n \n Results<\/jats:title>\n \n The proposed TabNet+GB+RF custom soft voting achieved significant, nonrandom performance over tuned base models with an\n F<\/jats:italic>\n 1<\/jats:sub>\n -score of 0.7111, area under the curve (AUC) of 0.8847 for Kenya. Conversely, for Uganda, the TabNet hybrid ensemble (GB+RF+ AB\u2192TabNet) was the best with AUC of 0.8862 and\n F<\/jats:italic>\n 1<\/jats:sub>\n -score of 0.6756, which was statistically insignificant. SHAP identified child\u2019s age and vitamin A1 as universal key predictors in both countries; others varied. In Kenya, time to water source, respondents\u2019 education, district lived previously, and current district were the most influential global features, respectively. In Uganda, Polio inactive, Pentavalent3, current district, and respondents\u2019 education were the most influential factors to defaulting, respectively.\n <\/jats:p>\n <\/jats:sec>\n \n Conclusions<\/jats:title>\n The proposed advanced ensemble learning models improved the classification of measles vaccine defaulters compared to individual models. The study recommends distinct country-tailored strategies. In Kenya, outreaches are recommended to areas with low literacy, distant water sources, and mobile populations. Conversely, in Uganda, the study recommends reminders to avoid missing preceding vaccines and outreaches to high-default districts.<\/jats:p>\n <\/jats:sec>","DOI":"10.2196\/95727","type":"journal-article","created":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T15:02:16Z","timestamp":1791903736000},"page":"e95727-e95727","source":"Crossref","is-referenced-by-count":0,"title":["Predicting Measles Vaccination Defaulting in Uganda and Kenya: Comparative Evaluation of Custom Soft Voting and TabNet Stacked Ensemble Architectures"],"prefix":"10.2196","volume":"5","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-5319-414X","authenticated-orcid":false,"given":"Bartha Alexandra","family":"Nantongo","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4731-2496","authenticated-orcid":false,"given":"Josephine","family":"Nabukenya","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2141-7940","authenticated-orcid":false,"given":"Peter","family":"Nabende","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1010","published-online":{"date-parts":[[2026,10,13]]},"reference":[{"issue":"1","key":"R1","doi-asserted-by":"publisher","DOI":"10.1186\/s13561-020-00262-3","article-title":"The effects of health expenditure on infant mortality in sub-Saharan Africa: evidence from panel data analysis","volume":"10","author":"Kiross","journal-title":"Health Econ Rev"},{"key":"R2","unstructured":"Uganda Demographic and Health Survey 2022. Uganda Bureau of Statistics; 2023. URL: https:\/\/www.ubos.org\/wp-content\/uploads\/publications\/UDHS-2022-Report.pdf [Accessed 11-09-2026]"},{"key":"R3","unstructured":"Kenya Demographic and Health Survey 2022. Kenya National Bureau of Statistics; 2023. URL: https:\/\/dhsprogram.com\/pubs\/pdf\/pr143\/pr143.pdf [Accessed 11-09-2026]"},{"key":"R4","doi-asserted-by":"crossref","unstructured":"Galadima AN Zulkefli NAM Said SM Ahmad N . Factors influencing childhood immunization uptake in Africa: A systematic review. Research Square. Preprint posted online on Jul 6, 2020. [doi: 10.21203\/rs.3.rs-39516\/v1]","DOI":"10.21203\/rs.3.rs-39516\/v1"},{"key":"R5","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijregi.2022.08.017","article-title":"Measles outbreak in Semuto Subcounty, Nakaseke District, Uganda, June-August 2021","volume":"5","author":"Nsubuga","journal-title":"IJID Reg"},{"issue":"3","key":"R6","doi-asserted-by":"publisher","first-page":"120","DOI":"10.17816\/PED83120-129","article-title":"Actual problems of measles","volume":"8","author":"Timchenko","journal-title":"Pediatr (SPb)"},{"issue":"3","key":"R7","doi-asserted-by":"publisher","DOI":"10.2196\/publichealth.9681","article-title":"Using predictive analytics to identify children at high risk of defaulting from a routine immunization program: Feasibility study","volume":"4","author":"Chandir","journal-title":"JMIR Public Health Surveill"},{"key":"R8","unstructured":"Uganda CO humanitarian annual situation report. UNICEF; 2018. URL: https:\/\/www.unicef.org\/media\/76021\/file\/Uganda-SitRep-Dec-2018.pdf [Accessed 21-09-2026]"},{"key":"R9","volume":"9","author":"Kizza","journal-title":"Uganda Public Health Bull"},{"issue":"10183","key":"R10","doi-asserted-by":"publisher","first-page":"1843","DOI":"10.1016\/S0140-6736(19)30226-0","article-title":"Mapping diphtheria-pertussis-tetanus vaccine coverage in Africa, 2000\u20132016: a spatial and temporal modelling study","volume":"393","author":"Mosser","journal-title":"The Lancet"},{"issue":"9","key":"R11","doi-asserted-by":"publisher","DOI":"10.2471\/blt.07.045328","article-title":"Use and abuse of rapid monitoring to assess coverage during mass vaccination campaigns","volume":"85","author":"Luman","journal-title":"Bull World Health Organ"},{"key":"R12","author":"Zziwa","journal-title":"Int J Public Health Res"},{"key":"R13","unstructured":"Dietz V Dietz-Chavez D . Vaccine Coverage Monitoring,\" in Vaccinology in Latin America: A Resource for Immunization Managers. Sabin Vaccine Institute; 2018:234-241. URL: https:\/\/www.sabin.org\/resources\/vaccinology-in-latin-america-a-resource-for-immunization-managers\/ [Accessed 11-09-2026]"},{"issue":"39","key":"R14","doi-asserted-by":"publisher","first-page":"6174","DOI":"10.1016\/j.vaccine.2020.05.058","article-title":"Combining cluster surveys to estimate vaccination coverage: experiences from Nigeria\u2019s multiple indicator cluster survey \/ national immunization coverage survey (MICS\/NICS), 2016-17","volume":"38","author":"Rhoda","journal-title":"Vaccine"},{"key":"R15","unstructured":"Wonodi C Stokes-Prindle C Aina M et al. Landscape analysis of routine immunization in Nigeria. Johns Hopkins Bloomberg School of Public Health; Jan 2012. [doi: 10.13140\/RG.2.2.16089.16483]"},{"issue":"12","key":"R16","doi-asserted-by":"publisher","first-page":"1080","DOI":"10.1093\/jpids\/piab073","article-title":"A risk prediction model to identify newborns at risk for missing early childhood vaccinations","volume":"10","author":"Oster","journal-title":"J Pediatric Infect Dis Soc"},{"key":"R17","author":"Mengesha","journal-title":"AIMS 2025 Workshop T2P"},{"issue":"2","key":"R18","doi-asserted-by":"publisher","first-page":"810","DOI":"10.51519\/journalisi.v6i2.689","article-title":"An artificial neural network model for predicting children at risk of defaulting from routine immunization in Nigeria","volume":"6","author":"Evwiekpaefe","journal-title":"journalisi"},{"issue":"4","key":"R19","doi-asserted-by":"publisher","DOI":"10.1093\/jamiaopen\/ooae132","article-title":"A retrospective cohort study on predicting infants at a risk of defaulting routine immunization in Uganda using machine learning models","volume":"7","author":"Nantongo","journal-title":"JAMIA Open"},{"issue":"8","key":"R20","doi-asserted-by":"publisher","first-page":"6679","DOI":"10.1609\/aaai.v35i8.16826","article-title":"TabNet: attentive interpretable tabular learning","volume":"35","author":"Arik","journal-title":"AAAI"},{"issue":"1","key":"R21","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-024-82282-1","article-title":"Enhancing predictive accuracy for urinary tract infections post-pediatric pyeloplasty with explainable AI: an ensemble TabNet approach","volume":"15","author":"Wang","journal-title":"Sci Rep"},{"issue":"5","key":"R22","doi-asserted-by":"publisher","first-page":"371","DOI":"10.1089\/big.2021.0257","article-title":"A comparative performance assessment of optimized multilevel ensemble learning model with existing classifier models","volume":"10","author":"Kumar","journal-title":"Big Data"},{"key":"R23","doi-asserted-by":"crossref","unstructured":"Verma A Mehta S . A comparative study of ensemble learning methods for classification in bioinformatics. Presented at: 2017 7th International Conference on Cloud Computing, Data Science & Engineering - Confluence (Confluence); Jan 12-13, 2017; Noida, India. [doi: 10.1109\/CONFLUENCE.2017.7943141]","DOI":"10.1109\/CONFLUENCE.2017.7943141"},{"key":"R24","doi-asserted-by":"crossref","unstructured":"Hansen ND Lioma C M\u00f8lbak K . Ensemble learned vaccination uptake prediction using web search queries. Presented at: CIKM \u201916: Proceedings of the 25th ACM International on Conference on Information and Knowledge Management; Oct 24-28, 2016; Indianapolis, IN. 2016.[doi: 10.1145\/2983323.2983882]","DOI":"10.1145\/2983323.2983882"},{"issue":"23","key":"R25","doi-asserted-by":"publisher","first-page":"12128","DOI":"10.3390\/app122312128","article-title":"Ensemble machine learning for monkeypox transmission time series forecasting","volume":"12","author":"Dada","journal-title":"Appl Sci"},{"key":"R26","unstructured":"Kunapuli G . Ensemble Methods for Machine Learning. Manning Publications; 2023. URL: https:\/\/www.manning.com\/books\/ensemble-methods-for-machine-learning [Accessed 11-09-2026]"},{"key":"R27","unstructured":"G\u00e9ron A . Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow. O\u2019Reilly Media; 2023. URL: https:\/\/www.oreilly.com\/library\/view\/hands-on-machine-learning\/9781492032632\/ [Accessed 11-09-2026]"},{"key":"R28","unstructured":"Ahishakiye E Wario R Mwangu W Taremwa D . Prediction of cervical cancer basing on risk factors using ensemble learning. Presented at: IST-Africa 2020 Conference; May 18-22, 2020; Kampala, Uganda. URL: https:\/\/ieeexplore.ieee.org\/document\/9144026 [Accessed 21-09-2026]"},{"key":"R29","unstructured":"Bishop CM . Pattern Recognition and Machine Learning. Springer; 2006. URL: https:\/\/link.springer.com\/book\/9780387310732 [Accessed 11-09-2026]"},{"issue":"2","key":"R30","doi-asserted-by":"publisher","first-page":"103","DOI":"10.3233\/HIS-2007-4204","article-title":"On diversity and accuracy of homogeneous and heterogeneous ensembles","volume":"4","author":"Bian","journal-title":"HIS"},{"key":"R31","unstructured":"Uganda demographic and health survey 2022. Uganda Bureau of Statistics. May 19, 2026. URL: https:\/\/microdata.ubos.org:7070\/index.php\/catalog\/81\/get-microdata [Accessed 08-06-2026]"},{"key":"R32","unstructured":"Demographic and health survey key findings. Uganda Bureau of Statistics (UBOS) and ICF; 2017. URL: https:\/\/dhsprogram.com\/pubs\/pdf\/FR333\/FR333.pdf [Accessed 21-09-2026]"},{"issue":"1","key":"R33","doi-asserted-by":"publisher","first-page":"118","DOI":"10.3329\/jhpn.v31i1.14756","article-title":"Factors influencing childhood immunization in Uganda","volume":"31","author":"Bbaale","journal-title":"J Health Popul Nutr"},{"key":"R34","unstructured":"Kembabazi B . A classification model leveraging electronic immunization records to predict child immunization completion: case study - Mukono health facility. Strathmore University; 2021. URL: https:\/\/su-plus.strathmore.edu\/server\/api\/core\/bitstreams\/1de6ebce-b5ec-49cd-a171-06c4fec18ac2\/content [Accessed 11-09-2026]"},{"issue":"1","key":"R35","doi-asserted-by":"publisher","DOI":"10.2196\/iproc.8559","article-title":"Using predictive analytics to prevent missed opportunities and achieve higher immunization coverage and timeliness","volume":"3","author":"Chandir","journal-title":"iProc"},{"issue":"1","key":"R36","doi-asserted-by":"publisher","first-page":"589","DOI":"10.32604\/cmc.2020.012507","article-title":"Smart healthcare using data-driven prediction of immunization defaulters in expanded program on immunization (EPI)","volume":"66","author":"Qazi","journal-title":"Comput Mater Contin"},{"issue":"1","key":"R37","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1177\/1367493508098380","article-title":"Factors predicting BCG immunization status in northern Nigeria: a behavioral-ecological perspective","volume":"13","author":"Babalola","journal-title":"J Child Health Care"},{"key":"R38","doi-asserted-by":"publisher","DOI":"10.1016\/j.csda.2020.107043","article-title":"A new correlation coefficient between categorical, ordinal and interval variables with Pearson characteristics","volume":"152","author":"Baak","journal-title":"Comput Stat Data Anal"},{"issue":"12","key":"R39","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0310839","article-title":"The impact of neglecting feature scaling in k-means clustering","volume":"19","author":"Wongoutong","journal-title":"PLOS One"},{"key":"R40","unstructured":"Brownlee J . Brownlee J , editor. Machine Learning Mastery With Python: Understand Your Data, Create Accurate Models, and Work Projects End-to-End. 4th ed. 2016. URL: https:\/\/share.google\/m9nVpTfvYKskVTlbX [Accessed 11-09-2026]"},{"key":"R41","unstructured":"Odilovich RN Dilmurod K . DATA PREPROCESSING TECHNIQUES IN MACHINE LEARNING. Research Bulletin; 2022:132-137."},{"issue":"4","key":"R42","doi-asserted-by":"publisher","first-page":"341","DOI":"10.4236\/jdaip.2020.84020","article-title":"Basic tenets of classification algorithms k-nearest-neighbor, support vector machine, random forest and neural network: a review","volume":"08","author":"Boateng","journal-title":"J Data Anal Inf Process"},{"key":"R43","unstructured":"Russell S Norvig P . Introduction. In: Artificial Intelligence: A Modern Approach. Prentice Hall; 2003. URL: https:\/\/dai.fmph.uniba.sk\/courses\/ICI\/References\/rn.chap1.pdf [Accessed 22-09-2026]"},{"key":"R44","unstructured":"Nti IK Nyarko-Boateng O Aning J . Performance of Machine Learning Algorithms with Different K Values in K-Fold Cross-Validation. I.J. Information Technology and Computer Science; 2021."},{"issue":"1","key":"R45","doi-asserted-by":"publisher","DOI":"10.1080\/08839514.2023.2176614","article-title":"A hybrid feature extraction and classification using Xception-RF for multiclass disease classification in plant leaves","volume":"37","author":"A","journal-title":"Appl Artif Intell"},{"issue":"1","key":"R46","doi-asserted-by":"publisher","first-page":"375","DOI":"10.32604\/cmc.2022.019167","article-title":"Improving routine immunization coverage through optimally designed predictive models","volume":"70","author":"Sameen","journal-title":"Comput Mater Contin"},{"key":"R47","doi-asserted-by":"publisher","DOI":"10.3389\/fdata.2025.1522578","article-title":"A stacked ensemble machine learning model for the prediction of pentavalent 3 vaccination dropout in East Africa","volume":"8","author":"Alemayehu","journal-title":"Front Big Data"},{"key":"R48","unstructured":"The DHS Program. URL: https:\/\/dhsprogram.com\/ [Accessed 14-09-2026]"}],"container-title":["JMIR AI"],"original-title":[],"language":"en","deposited":{"date-parts":[[2026,10,13]],"date-time":"2026-10-13T15:02:18Z","timestamp":1791903738000},"score":1,"resource":{"primary":{"URL":"https:\/\/ai.jmir.org\/2026\/1\/e95727"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,13]]},"references-count":48,"URL":"https:\/\/doi.org\/10.2196\/95727","relation":{},"ISSN":["2817-1705"],"issn-type":[{"value":"2817-1705","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,13]]},"article-number":"v5i1e95727"}}