A machine learning‐based multiple imputation method for the health and aging brain among latino elders study

Background The Health and Aging Brain among Latino Elders Study (HABLE) project seeks to understand the biological, social environmental factors that impact brain aging diverse communities. HABLE, like many other NIH funded data-sharing projects, has important data assets for various uses, including social, behavioral data, multiple flow pathways. A common methodological issue collection in such a large survey-based, behavior, behavioral, or epidemiology study is missing data. It impossible achieve machine learning (ML) if contains values. Therefore, developing new imputation methodology become very urgent task HABLE. Method In this study, we developed ML-Based Multiple Imputation method. First, built four models (Support Vector Machine (SVM), Random Forest (RF), eXtreme Gradient Boosting (XGB) lasso elastic-net regularized Generalized Linear Model (GLMNET)). Then adapted above methods imputation. Lastly, evaluated our method comparison with lowest of detection (LOD), random drawing from lef-censored normal distribution (ND), k-nearest neighbors (kNN), local least squares (LLS), forest Singular value decomposition (SVD), Bayesian principal component analysis (BPCA). Result Our results show can have lower Normalized Root Mean Square Error (NRMSE) higer prediciton performance learning. Conclusion choice handling significant on accompanying statistical analyses alzheimer’s disease. Learning-Based performs well disease models. also be applied analysis.