Fig. 2 | Scientific Reports

Fig. 2

From: Development of an early prediction model for vomiting during hemodialysis using LASSO regression and Boruta feature selection

Fig. 2

While LASSO is used for significant variable selection, the Boruta method is employed for feature selection. (A) LASSO coefficient profiles of the clinical characteristics. (B) The optimal penalization coefficient lambda was obtained using tenfold cross-validation in LASSO. The graphic shows the lambda value of the most minor mean square error. The most minor absolute shrinkage and selection operator is known as LASSO. (C) The function selected by Boruta. The blue boxplots correspond to the minimum, mean and maximum Z-scores of the shaded attributes. Z-scores clearly distinguish between important and non-important attributes. Red and green represent attributes that Boruta chooses to reject and confirm, respectively.

Back to article page