Table 2 Main model: XGBoost model with all available features included.

From: Identifying top ten predictors of type 2 diabetes through machine learning analysis of UK Biobank data

Confusion matrix for main model

Confusion matrix for reduced model

 

Truth

 

Truth

1

0

1

0

Prediction

1

1554

5902

Prediction

1

1419

5976

0

941

81,259

0

1076

81,185

  1. Reduced model: XGBoost model with only the 10 most influential features based on Shap values. These matrices summarize the performance of each model in classifying instances as positive (1) or negative (0).