VdaBSC: A Novel Vulnerability Detection Approach for Blockchain Smart Contract by Dynamic Analysis

Table 7

Comparative analysis of proposed model with state-of-the-art deep learning techniques.

Model	Accuracy (%)	Precision (%)	Recall (%)	F1-score (%)	Computational time (s)
HAM	88.2	87.4	88.9	88.1	12
BiLSTM-ATT	89.0	88.3	89.5	88.9	15
DR-GCN	87.5	86.8	88.1	87.4	20
TMP	86.9	86.2	87.5	86.8	18
LSTM	85.7	85.0	86.3	85.6	14
Vanilla-RNN	83.4	82.7	84.0	83.3	10
Proposed model	96.5	96.0	95.8	95.9	8