Table 6.

Prediction result of toxicity (CC₅₀) using SMILES to ECFP

Embedding method	Classification	Accuracy	Precision	Recall	F1
word2vec	LR	0.879	0.913	0.929	0.921
	LDA	0.874	0.916	0.918	0.917
	KNN	0.886	0.911	0.942	0.926
	CART	0.858	0.915	0.897	0.906
	NB	0.710	0.833	0.772	0.801
	SVM	0.893	0.893	0.977	0.933
	XGBoost	0.912	0.918	0.971	0.943
	RDForest	0.907	0.911	0.972	0.940
Ising-word2vec	LR	0.888	0.914	0.941	0.927
	LDA	0.883	0.921	0.925	0.923
	KNN	0.884	0.910	0.940	0.924
	CART	0.861	0.912	0.904	0.908
	NB	0.710	0.832	0.774	0.802
	SVM	0.889	0.890	0.975	0.930
	XGBoost	0.912	0.916	0.971	0.943
	RDForest	0.904	0.907	0.973	0.939