Skip to main content
Fig. 6 | BMC Psychology

Fig. 6

From: Predicting depression and unravelling its heterogeneous influences in middle-aged and older people populations: a machine learning approach

Fig. 6

Differences in Predictor Importance Between CNN and RF Models (6 A: Feature importance in the CNN model computed using SHAP; 6B: Feature importance in the RF model computed using SHAP, where Class 1 represents depression and Class 0 represents non-depression; 6 C: Feature importance in the RF model computed using impurity-based methods.)

Back to article page

BMC Psychology

ISSN: 2050-7283

Contact us