Fig. 6: The scheme of the data analysis and construction of cognitive age models.
From: Cognitive aging clock based on Explainable artificial intelligence
The scheme includes 11 blocks that briefly describe the steps of the pipeline: Data collection (Section “Study design”), Data preprocessing (Section “Data preprocessing”), Feature subsets (Section “Feature subsets”), Building regression models (Section “Machine-learning models”), Model selection (Section “Model selection and evaluation”), Best feature selection (Section “Feature selection”), Model training (Section “Feature selection”), Bias correction (Section “Bias correction”), Model assessment (Section “Model selection and evaluation”), Global explanation (Section “Model explainability”), Local explanation (Section “Model explainability”).
