Predictive model and risk analysis for coronary heart disease in people living with HIV using machine learning

Liu, Zengjing; Meng, Zhihao; Wei, Di; Qin, Yuan; Lv, Yu; Xie, Luman; Qiu, Hong; Xie, Bo; Li, Lanxiang; Wei, Xihua; Zhang, Die; Liang, Boying; Li, Wen; Qin, Shanfang; Yan, Tengyue; Meng, Qiuxia; Wei, Huilin; Jiang, Guiyang; Su, Lingsong; Jiang, Nili; Zhang, Kai; Lv, Jiannan; Hu, Yanling

doi:10.1186/s12911-024-02511-5

Fig. 4

From: Predictive model and risk analysis for coronary heart disease in people living with HIV using machine learning

SHAP dependence plot of the LightGBM model, illustrating the influence of a single variable on the prediction. The blue points in the plot represent the SHAP values for this feature across different data points, with the horizontal position indicating the feature’s contribution to the predictive outcome. SHAP values greater than zero suggest an increased risk of CHD in PLHIV. The vertical axis represents the actual values of the feature, and the distribution of points reveals the relationship between the feature values and the risk of CHD. Glu, glucose; Scr, serum creatinine; I-Bil, indirect bilirubin; AMY, amylase; SUA, serum uric acid

Back to article page

Fig. 4

BMC Medical Informatics and Decision Making

Contact us