A Robust Text Classifier Based on Denoising Deep Neural Network in the Analysis of Big Data

Table 4

Text classification performance with different models using 20-Newsgroup dataset.

	Classifier	Noise factor
		0.00	0.01	0.02	0.03	0.04	0.05
Accuracy	NB	0.7506	0.7274	0.6895	0.6678	0.5887	0.4633
	KNN	0.6136	0.6161	0.6213	0.6142	0.6043	0.5978
	SVM	0.7598	0.7527	0.7294	0.6968	0.6652	0.6453
	DBN	0.7235	0.7207	0.7041	0.6849	0.6562	0.6252
	DDNN	0.7536	0.7561	0.7550	0.7542	0.7443	0.7378
Recall	NB	0.7483	0.6693	0.5053	0.3526	0.2613	0.2027
	KNN	0.5959	0.6000	0.6070	0.6034	0.5939	0.5820
	SVM	0.7525	0.7415	0.6966	0.6094	0.4891	0.3833
	DBN	0.7149	0.7120	0.6990	0.6826	0.6439	0.6250
	DDNN	0.7459	0.7500	0.7549	0.7534	0.7439	0.7320
-score	NB	0.7494	0.6971	0.5832	0.4615	0.3619	0.2820
	KNN	0.6046	0.6079	0.6141	0.6088	0.5991	0.5898
	SVM	0.7561	0.7471	0.7126	0.6502	0.5637	0.4809
	DBN	0.7192	0.7163	0.7015	0.6837	0.6500	0.6251
	DDNN	0.7497	0.7530	0.7549	0.7538	0.7441	0.7349