Fig. 5From: Development and validation of a machine learning model to predict time to renal replacement therapy in patients with chronic kidney diseaseRelationship between predicted and measured values for the conventional eGFR time series method.Predictions tended to be very large, with a maximum predicted value of 73,549, which is considered an outlier, and for readability, the Y-axis is shown with a range up to 50,000Back to article page