Multimodal Feature Fusion Based Hypergraph Learning Model

Table 3

Classification performance of modal combinations.

Datasets	Modal combinations	Incidence matrix extension				Incidence matrix extension & Laplacian matrix fusion
		Accuracy	Precision	Recall	F1	Accuracy	Precision	Recall	F1
Cat & Dog	Poor + Poor	0.687	0.662	0.674	0.667	0.722	0.718	0.720	0.718
	PHA + SIFT
	Poor + Good	0.919	0.920	0.919	0.919	0.928	0.926	0.930	0.928
	PHA + RVGG
	Good + Good	0.975	0.975	0.977	0.976	0.986	0.986	0.986	0.986
	RVGG + HVGG
Ctrip	Poor + Poor	0.619	0.620	0.619	0.619	0.623	0.621	0.623	0.621
	Jaccard + TF-IDF
	Good + Poor	0.646	0.821	0.613	0.701	0.648	0.837	0.624	0.711
	TF-IDF + Doc2vec
	Good + Good	0.659	0.878	0.709	0.785	0.722	0.718	0.720	0.718
	Doc2vec + word2vec