Table 6.

Performance of different prediction models for rejection and mortality.

Prediction models				Precision	Recall		F1-score		AUROCa	AUPRCb
Rejection
	At 1 year
		XGBoostc	0.688		0.726	0.691		0.641		0.576
		LRd	0.698		0.737	0.679		0.648		0.576
		SVMe	0.531		0.728	0.614		0.485		0.614
		RFf	0.695		0.735	0.677		0.664		0.575
		SGDg	0.641		0.611	0.623		0.547		0.592
		MLPh	0.662		0.712	0.668		0.627		0.578
		AdaBoosti	0.699		0.735	0.696		0.648		0.576
		NNj	0.610		0.699	0.629		0.504		0.604
	At 3 years
		XGBoost	0.717		0.768	0.728		0.695		0.739
		LR	0.709		0.779	0.711		0.692		0.737
		SVM	0.617		0.785	0.691		0.480		0.663
		RF	0.724		0.785	0.707		0.706		0.738
		SGD	0.680		0.677	0.679		0.523		0.668
		MLP	0.697		0.766	0.712		0.675		0.733
		AdaBoost	0.717		0.769	0.728		0.703		0.734
		NN	0.673		0.780	0.694		0.491		0.664
	At 5 years
		XGBoost	0.873		0.915	0.881		0.697		0.888
		LR	0.841		0.916	0.877		0.685		0.885
		SVM	0.841		0.917	0.877		0.462		0.841
		RF	0.841		0.917	0.877		0.676		0.882
		SGD	0.853		0.816	0.833		0.526		0.851
		MLP	0.847		0.905	0.873		0.667		0.882
		AdaBoost	0.866		0.911	0.880		0.705		0.887
		NN	0.853		0.915	0.877		0.484		0.847
Mortality
	At 1 year
		XGBoost	0.878		0.926	0.896		0.663		0.838
		LR	0.899		0.929	0.895		0.669		0.835
		SVM	0.863		0.929	0.895		0.502		0.868
		RF	0.863		0.929	0.895		0.697		0.834
		SGD	0.875		0.912	0.891		0.534		0.859
		MLP	0.887		0.928	0.897		0.652		0.837
		AdaBoost	0.886		0.926	0.898		0.667		0.838
		NN	0.863		0.927	0.894		0.493		0.868
	At 3 years
		XGBoost	0.725		0.745	0.729		0.737		0.567
		LR	0.709		0.739	0.699		0.719		0.566
		SVM	0.626		0.722	0.607		0.574		0.584
		RF	0.718		0.745	0.706		0.758		0.569
		SGD	0.646		0.596	0.614		0.564		0.584
		MLP	0.707		0.735	0.707		0.711		0.567
		AdaBoost	0.720		0.744	0.720		0.738		0.565
		NN	0.603		0.677	0.623		0.503		0.600
	At 5 years
		XGBoost	0.688		0.690	0.689		0.748		0.575
		LR	0.668		0.671	0.669		0.718		0.559
		SVM	0.577		0.588	0.555		0.613		0.530
		RF	0.717		0.718	0.717		0.763		0.574
		SGD	0.599		0.604	0.600		0.596		0.521
		MLP	0.636		0.638	0.622		0.683		0.550
		AdaBoost	0.692		0.692	0.692		0.735		0.562
		NN	0.508		0.534	0.501		0.517		0.514

^aAUROC: area under the receiver operating characteristic curve.

^bAUPRC: area under the precision-recall curve.

^cXGBoost: extreme gradient boosting.

^dLR: logistic regression.

^eSVM: support vector machine.

^fRF: random forest.

^gSGD: stochastic gradient descent.

^hMLP: multilayer perceptron.

ⁱAdaBoost: adaptive boosting.

^jNN: neural network.