Table 2.

Performance comparison of different split char lengths in SMILES with SC

Embedding model	Splitted char length	Score	LR	LDA	KNN	CART	NB	SVM	XG-Boost	RD-Forest
Ising-word2vec	4	Accuracy	0.790	0.788	0.794	0.736	0.706	0.854	0.849	0.834
		Precision	0.698	0.701	0.655	0.597	0.544	0.832	0.827	0.883
		F1	0.668	0.661	0.718	0.606	0.599	0.760	0.750	0.696
		Recall	0.641	0.627	0.796	0.616	0.666	0.699	0.687	0.575
	3	Accuracy	0.776	0.771	0.802	0.728	0.686	0.844	0.845	0.838
		Precision	0.691	0.687	0.679	0.595	0.531	0.826	0.819	0.882
		F1	0.654	0.641	0.732	0.610	0.584	0.750	0.753	0.716
		Recall	0.621	0.602	0.795	0.626	0.650	0.687	0.697	0.603
word2vec	4	Accuracy	0.786	0.784	0.798	0.721	0.709	0.851	0.848	0.835
		Precision	0.692	0.693	0.665	0.573	0.548	0.831	0.826	0.887
		F1	0.661	0.655	0.720	0.587	0.603	0.754	0.749	0.696
		Recall	0.634	0.621	0.786	0.603	0.671	0.690	0.685	0.573
	3	Accuracy	0.777	0.772	0.799	0.736	0.693	0.844	0.847	0.839
		Precision	0.690	0.689	0.676	0.610	0.541	0.822	0.826	0.881
		F1	0.657	0.644	0.726	0.617	0.586	0.750	0.753	0.720
		Recall	0.627	0.604	0.785	0.624	0.639	0.690	0.698	0.609