Fig. 7 | Scientific Reports

Fig. 7

From: Identification of UBE2N as a biomarker of Alzheimer’s disease by combining WGCNA with machine learning algorithms

Fig. 7

Validating and interpreting machine learning models. (A) Specific values of AUC in the XGBoost model. (B) Specific values of P-R in the XGBoost model. (C) The importance matrix shows the contribution of each feature gene in the XGBoost model. (D) The SHAP summary plot shows the contribution of each feature to the XGBoost model. (D, E) The left side of the figure shows the predicted results of the LIME. The ten variables with the greatest impact on normal or AD onset are listed on the right side. The length of each feature bar indicates the importance of the corresponding feature in the prediction. (G, H) Predicted probabilities of normal and AD onset based on SHAP force maps, respectively. Red and blue bars represent increased and decreased likelihood of AD, respectively.

Back to article page