Fig. 2
From: Clinical data mining: challenges, opportunities, and recommendations for translational applications
Clinical data mining studies analyze workflows. The analytic workflows for developing a model consist of eight steps, including phase 1: problem definition (show in purple); phase 2: data collection (show in orange); phase 3: data processing (show in light green); phase 4: feature analysis (show in blue); phase 5: model developing (show in light yellow); phase 6: model validation and visualization (show in dark yellow); phase 7:model evaluation (show in dark green); phase 8: model application and upgrades (show in pink)