Skip to main content

Table 7 Model performance comparison—metric scores

From: Short-term stock market price trend prediction using a comprehensive deep learning system

Model F1 score Binary accuracy TPR (recall) TNR (specificity) FPR (fall-out) FNR (miss rate)
LR 0.90 0.90 0.92 0.88 0.08 0.12
SVM 0.90 0.90 0.92 0.88 0.08 0.12
NB 0.89 0.89 0.90 0.87 0.10 0.13
MLP (Single hidden layer) 0.90 0.90 0.92 0.88 0.08 0.12
MLP (3 hidden layers) 0.90 0.90 0.92 0.87 0.08 0.13
MLP (5 hidden layers) 0.90 0.90 0.92 0.88 0.08 0.12
RAF 0.88 0.88 0.94 0.81 0.06 0.19
Proposed model 0.93 0.93 0.96 0.90 0.04 0.10