Table 2 Comparative evaluation of machine learning models using standard performance metrics: accuracy, precision, recall, F1 score, and MCC. The table presents quantitative measurements ranging from 0.746 to 1.000 across different architectures, demonstrating varying levels of predictive performance and classification effectiveness in binary classification tasks.
From: Deep learning decodes species-specific codon usage signatures in Brassica from coding sequences
Model Name | Accuracy | Precision | Recall | F1 Score | MCC |
|---|---|---|---|---|---|
Deep Belief Neural Network | 0.99995 | 0.99994 | 0.99994 | 0.99994 | 0.99993 |
Multilayer Perceptron Neural Network | 1 | 1 | 1 | 1 | 1 |
Deep neural network (DNN) with L2 regularization and dropout | 0.99982 | 0.99981 | 0.99981 | 0.99981 | 0.99975 |
Leaky ReLU Neural Network | 0.99998 | 0.99998 | 0.99998 | 0.99998 | 0.99997 |
Shallow Neural Network | 0.99995 | 0.99993 | 0.99995 | 0.99994 | 0.99994 |
Dropout Neural Network | 0.99998 | 0.99998 | 0.99997 | 0.99998 | 0.99998 |
Radial Basis Function Neural Network | 0.7464 | 0.85665 | 0.70124 | 0.74178 | 0.6673 |
