Fig. 4From: Importance of medical data preprocessing in predictive modeling and risk factor discovery for the frailty syndromeModeling and Evaluation Procedure. This image shows the general modeling and evaluation procedure. Firstly, models are built using the 5 different obtained imputed data sets. Secondly, the models are evaluated in a cross-fold validation setup. Then the resulting performance measure values (e.g. accuracy, sensitivity, specificity) can be comparedBack to article page