Regency EF
WITHDRAWN: A Comparison of Linear Mixed Effect Model and REEM Tree for Prediction of Cognitive Decline (304002)
Hansapani Rodrigo, The University of Texas Rio Grande ValleyKristina P Vatcheva, The University of Texas Rio Grande Valley
Keywords: mixed effects models, REEM trees
Cognitive decline is common with ageing, but other risk factors may influence this process. Cognitive decline can have profound implications for an individual’s well-being and its prediction and early detection can prevent and improve their lives and decrease hospitalization cost. We compared the predictive ability of linear mixed effects model (LMEM) and REEM tree on cognitive decline. Data from five waves of the English Longitudinal Study of Aging (ELSA) were analyzed. The outcome variable was index of memory function and the predictors considered in the models were age, sex, diabetes history, and other known risk factors. REEM tree and two LMEMs, with predictors selected by REEM and with all predictors, with random intercept and slope for time, were fitted on training data. Models' prediction abilities were evaluated on test data using root mean squared error (RMSE). Data were unbalanced and comprised of 12, 212 participants with a total of 42, 560 records. LMEMs assumptions were satisfied. Based on the results, both liner mixed effects models resulted with better prediction performance compared to the fitted REEM tree (RMSE=3.57, RMSE=3.60 vs. RMSE=3.67, respectively).