Semantic Sentiment Classification for COVID-19 Tweets Using Universal Sentence Encoder

Table 6

Proposed model vs. CNN classifier based on TF-IDF.

Epochs#	Proposed model		CNN model
	Accuracy	Loss	Accuracy	Loss
10	77.742%	0.8005	72.16%	1.1315
20	77.982%	0.8006	70.70%	1.4414
30	77.902%	0.7993	69.87%	1.6880
40	77.982%	0.7994	69.09%	2.2059
50	77.9823	0.7987	69.55%	2.4477
60	78.0624	0.7991	69.23%	2.8105