Exploring Security Vulnerabilities of Deep Learning Models by Adversarial Attacks

Table 3

Accuracy of deep learning models on the Yelp review dataset.

Model	LSTM	BiLSTM
Accuracy	95.6934%	95.647%