2026-10-11T03:52:31Z https://api.figshare.com/v2/oai

oai:figshare.com:article/34025799 2026-09-29T17:25:37Z category_7 category_734 category_931 portal_5 item_type_1 month_year_09_2026

<p>McNemar’s test (CHIT vs. MICE) with 95% Wilson CIs — Heart Disease Dataset (n_test≈205).</p> Kurban Kotan (25138206) Serdar Kırışoğlu (25138209) Medicine Biological Sciences not elsewhere classified Information Systems not elsewhere classified wilson score method present work addresses health informatics pipelines gaussian naïve bayes full mcnemar results driven missingness patterns competing methods degrade chronic kidney disease 38 percentage points configuration mean accuracy classifier accuracy differences accuracy remains stable regression training set appropriate imputation strategy results provide mechanism perfect downstream classification heart disease dataset div >< p reported accuracy values chit achieves near missing data mechanism mcar ), missing logistic regression multiple imputation imputation selection hdd ), first mechanism taken together stratified comparison statistical significance standard deviation specific guidance scarce features record prioritisation overall rates mped ). iterative enrichment fig 2 experimental platforms decision tree corrected chi confidence intervals computed using based filling 96 ]. 80 ]. <p>Green = p < 0.001(***), Yellow = p < 0.05(*). All eight classifiers across MCAR, MAR, and MNAR.</p> <p>(PNG)</p> 2026-09-29T17:25:35Z Image Figure 10.1371/journal.pone.0359577.s001 https://figshare.com/articles/figure/_p_McNemar_s_test_CHIT_vs_MICE_with_95_Wilson_CIs_Heart_Disease_Dataset_n_test_205_p_/34025799 CC BY 4.0