Table 2 Performance metrics for feature selection using recursive feature elimination (RFE) in predicting depression.

From: An interpretable machine learning model predicts the interactive and cumulative risks of different environmental chemical exposures on depression

Variables

RMSE

R squared

MAE

RMSE SD

R-squared SD

MAE SD

6

0.2874

0.01506

0.1645

0.0054

0.004331

0.002355

8

0.2858

0.01748

0.1648

0.005382

0.00495

0.002105

10

0.2861

0.01827

0.166

0.005385

0.004347

0.002106

84

0.2831

0.03409

0.1663

0.005243

0.003852

0.002184

  1. Outer resampling method: Bootstrapped (25 reps). This table presents the performance metrics for different subsets of variables selected using recursive feature elimination (RFE). The RFE process was employed to identify the most significant features for predicting depression using the random forest (RF) model. The metrics include Root Mean Squared Error (RMSE), R squared, Mean Absolute Error (MAE), and their standard deviations (SD). The goal of RFE in this context is to enhance the predictive accuracy and interpretability of the RF model by identifying the most relevant features among the ECMs for predicting depression. The key features identified were Total Arsenic, Arsenous Acid, Cadmium, Cesium, 2,5-Dichlorophenol, and 1-Hydroxynaphthalene.
Back to article page