Table 2 Mean Scores for the best performing classifier in a 5-fold cross-validation of our dataset

From: Predicting adverse events for risk stratification of chemotherapy based stem cell mobilization in multiple myeloma

Target

Metric

Classifier

Mean score

MRI

   
 

Accuracy

Random Forest Classifier

0.96

 

ROC-AUC

XGBoost Classifier

1.0

 

MCC

XGBoost Classifier

0.78

Fever

   
 

Accuracy

Gradient Boost Classifier

0.9

 

ROC-AUC

Logistic Regression

0.67

 

MCC

Logistic Regression

0.4

Transfusion

   
 

Accuracy

Random Forest Classifier

0.9

 

ROC-AUC

Random Forest Classifier

0.91

 

MCC

TabPFN Classifier

0.48

Any AE

   
 

Accuracy

XGBoost Classifier

0.79

 

ROC-AUC

Logistic Regression

0.81

 

MCC

XGBoost Classifier

0.52

  1. MRI: prediction of supportive IV Fluid due to mild renal impairment; fever: neutropenic fever; transfusion needs; any AE: composite Endpoint of occurrence of any of the above adverse events.
  2. Hyperparameter-tuning was performed within nested cross-validation.